We the Citizens™ — Terms of Service

The Short Version

This summary is not part of the contract and does not change anything below it. Where the summary and the full terms differ, the full terms control. We wrote it because a movement built on honesty should not hide its own contract behind twelve thousand words. Each section below also opens with its own one-line summary, on the same footing (§1.8).

What you're agreeing to

You will not upload illegal or abusive content, and you will not post sexual content involving minors, non-consensual intimate imagery, threats, doxxing, or malware. (§8)
You will not make false statements of fact about real people. Opinion, criticism, and satire about public figures are the point of this platform and are welcome. Fabricated quotes, doctored documents, and knowingly false accusations are not. (§11)
You will be honest with evidence. If you knowingly falsify data — fabricate a quote, doctor a document, pass off synthetic media as real — you agree to pay us $400,000 per incident. Honest mistakes and typos are expressly excluded, and you can correct anything you got wrong without penalty. (§10)
You own the rights to what you upload, or you have permission to upload it. (§12)
You take responsibility for your own decisions. Nothing here is legal, financial, medical, or election-law advice, and our AI can be wrong. (§18, §21)

What we do not promise

We do not promise the Service is correct, available, secure, or that anything you publish will be read or acted on by anyone. It is provided as is. (§21)
We work to make content accurate, we use AI, AI makes mistakes, and we take no liability for the result. Please tell us when we get something wrong — there is a real address for it at §30. (§18.12, §21.9, §13.12)
WE TRAIN AI ON WHAT YOU PUBLISH — ours, and other companies' models. You grant that licence at §7.2A, it runs only while your separate consent under the Privacy Policy stands, and it is permanent for anything already trained, because a model cannot be untrained. Withdrawing consent stops future training. Content that never leaves your own machine is never trained on by anyone. (§7.2A, §18.6, Privacy Policy §11)
AI can be false, incomplete, misleading, or simply out of date about recent events. Check anything that matters before you rely on it. (§18.3, §18.13)
We do not promise any political or legislative outcome. (§21.5)
We cannot recover your local files. They live on your computer, not ours. Keep your own backups. (§5.2)
We cannot un-publish what you send to a federated network. Once it has spread to machines we do not run, deleting our copy does not delete theirs — and neither you nor we can call it back. (§5.4, §5.7, §19.3)
We are not a nonprofit, and nothing you pay us is tax-deductible. (§17.12)
Some of what these Terms describe is not built yet. Payments, WeTheCitizens.tv, and publishing to a federated network are written here in advance so the rules exist before the feature does. We are not promising to build any of them. (§3.7)

Free speech, and moderation

We hold free expression as a core commitment and we write it into this contract rather than only into our marketing. Precisely because we permit a wide range of speech, we also reserve the discretion to remove content and to suspend accounts — including accounts running coordinated narrative control on behalf of a government, an intelligence service, or political money. We tell you how that discretion is constrained: strikes expire, notices name the specific rule and the specific content, and you get a real appeal window with a deadline that runs against us as well as you. (§8, §9, §13)

Money

You can hold more than one paid subscription at the same time, and cancelling one does not cancel the others. Each one is a separate contract. (§15.2, §15.19)
Subscriptions renew automatically until you cancel. You can cancel any time, from inside the product, in no more steps than it took to sign up. (§15.3, §15.4)
We never pre-check a recurring box. Recurring requires you to actively choose it. (§15.6)
Deleting your account does not cancel your subscriptions. Cancel them first. (§15.20)
We do not give money back just because you ask, and a donation is a gift to support the work rather than money we hold for you — but the refund rights your own law gives you still stand, and nothing here switches them off. (§15.10, §15.24, §17.11, §15.21)
If you buy a film on WeTheCitizens.tv, you are buying a licence to watch it, not a copy — and it can go away if our rights to it end. (§16.2, §16.11)

If you have a problem with us

Tell us first. You agree to file a complaint and let our process run (§13.13, §25.11), then to send a Notice of Dispute (§25.1), then to attend mediation in King County, Washington (§25.10) — before starting any other proceeding. Those steps have real deadlines, they are all deemed satisfied after 120 days whatever stage they have reached, we attend in person too (§25.16), we pay the mediator, and they do not apply to small-claims court, to the EU complaint routes, or to your right to complain to a regulator (§25.13, §25.17).

Getting out

Export your data any time (§20.4), delete your account any time (§27.2, §27.9), and take your local files with you — they were always yours (§5.1).

Where this contract lives

You are contracting with ACT 3 AI, Inc., a Washington corporation, doing business as We the Citizens (§1.1). Washington law governs and King County, Washington is the venue (§26.1, §26.2) — except that if you are a consumer in the EU, the EEA, the UK, Switzerland, Norway, or Iceland, your own country's law and your own country's courts apply instead (§26.11). We do not try to drag you to Washington.

§		§
1	The Agreement and Acceptance	17	Donations, Fundraising, and Campaign Finance
2	Definitions	18	AI Features and AI Output
3	Scope — One Product, Many Front Doors	19	Third Parties, Federation, and Integrations
4	Eligibility, Accounts, and Identity	20	Privacy and Data Handling
5	The Local-First Split	21	Disclaimers of Warranty
6	License to Use the Service	22	Limitation of Liability
7	User Content and the License to Us	23	Release and Covenant Not to Sue
8	Acceptable Use	24	Indemnification
9	Inauthentic Accounts and Narrative Control	25	Dispute Resolution
10	Data Accuracy and Integrity	26	Governing Law, Venue, International
11	Defamation and Statements About Real People	27	Term, Suspension, Termination, Survival
12	Intellectual Property, Infringement, DMCA	28	Changes to These Terms
13	Moderation, Enforcement, and Appeals	29	General Provisions
14	Our Legal Role — Safe Harbors	30	Contact and Legal Notices
15	Payment, Credit Cards, and Subscriptions
16	Streaming, Pay-Per-View, and Rentals

1. The Agreement and Acceptance

In short: This is a contract between you and us. You accept it by using the hosted parts of the Service. Running the app purely on your own machine does not require it.

1.1 The parties. These Terms of Service (the "Terms") are a binding contract between ACT 3 AI, Inc., a corporation organized under the laws of the State of Washington, doing business as We the Citizens ("We the Citizens", "we", "us", "our"), and you, the individual using the Service ("you", and where the context is about a person using the movement, a "Citizen"). "We the Citizens" is used throughout to refer to both the company and the product.

ACT 3 AI, Inc. also operates other products under other brands, including our sibling social network. "We the Citizens" is a brand and assumed name of that one company, not a second or separate corporation, and you are contracting with ACT 3 AI, Inc. whichever of our properties in §3.1 or §3.2 you arrived through. Our Privacy Policy identifies the other brands by name, because a person is entitled to know which legal person holds their data (Privacy Policy §2.1a).

1.2 Acceptance. You accept these Terms by doing any of the following: creating an account; running the Local App in a way that connects to the Hosted Service; purchasing any paid item; or accessing any of the properties listed in §3.1. Purely local use of the Local App, with no connection to the Hosted Service, requires no acceptance of these Terms — that use is governed only by the open-source license on the code (§6.6).

1.3 Capacity and authority. You represent that you are a natural person of legal capacity, at least 18 years old (§4.1). If you accept these Terms on behalf of a community, organization, campaign, or employer, you represent that you have authority to bind it, and "you" includes that entity. If you are a government entity, or an official acting in an official capacity, read §29.14 first — parts of these Terms may not lawfully apply to you, and we would rather say so than pretend otherwise.

1.4 The summaries are not the contract. "The Short Version" above, and the "In short" line at the head of each section below, are plain-English summaries provided for readability. They are not binding and do not modify these Terms. Where a summary and the full Terms differ, the full Terms control.

1.5 Documents incorporated by reference. The following are separate documents and are incorporated into these Terms:

the Privacy Policy;
the Content Moderation Policy and Community Guidelines;
the DMCA and Copyright Policy (§12.7–§12.12);
the Accuracy and Corrections Policy — how we handle a mistake you report to us under §13.12, what we do about it, and what we publish when we correct something;
the open-source license on our source code (but see §6.6 — it governs the code and nothing else); and
any product-specific terms presented to you for a particular feature (§3.5).

1.6 Order of precedence. If these documents conflict, the order of precedence is: (a) product-specific terms you expressly accepted for a particular feature, but only as to that feature; (b) these Terms; (c) the Privacy Policy, except that the Privacy Policy controls on any question of what we do with personal data; (d) the Content Moderation Policy, the Accuracy and Corrections Policy, and Community Guidelines. The open-source license is not ranked in this list because it governs a different subject matter entirely (§6.6).

1.7 Last modified and versions. These Terms take effect on the last-modified date shown at the bottom. We keep a public version history and change log (§28.4).

1.8 Every section has a summary, and none of them is binding. Each section of these Terms opens with a short "In short" line. We do that because a contract nobody reads is a contract nobody consented to, and because the summary is most useful where the reader actually is rather than only at the top. Those lines are a reading aid. They are not part of the operative terms, they cannot create, enlarge, or reduce any right or obligation, and where one differs from the section it heads, the section controls (§1.4).

1.9 The Accuracy and Corrections Policy. (Folded into §1.5.) Because needs.txt requires us both to ask you to report mistakes and to take no liability for them, we split the two: these Terms carry the disclaimer, the no-liability statement, and the address (§18.12, §21.9, §30), and the separate Accuracy and Corrections Policy carries whatever we commit to doing about a report. Keeping the commitment out of the contract is deliberate — it lets us improve the process without amending your contract, and it stops a process description from being read as a promise of a result (§13.12).

2. Definitions

In short: The words we use, defined. The one that surprises people is "board of directors" — for us it means an advisory group with no authority (§2.8).

2.1 The product. "Service" means everything we offer under these Terms, taken together. "Site" means our websites, including the front doors listed in §3.2. "Local App" means the application you run on your own computer, where your files stay on your own machine. "Hosted Service" means the optional server-side service we operate for accounts, publishing, community aggregation, payments, and AI features. "Federated Network" means relays, media servers, instances, and other systems operated by third parties to which content may be published, including any protocol-compatible network.

2.2 People and places. "User" and "Citizen" mean an individual using the Service. "Account" means your registered identity on the Hosted Service. "Community" means a group of Citizens organized inside the Service, and "Area" means a geographic or topical subdivision within the layered content hierarchy (earth, country, state, community, area).

2.3 What you make. "User Content" means anything you submit, upload, publish, or transmit through the Service, including text, images, audio, video, files, and metadata. "Evidence" means a fact, figure, quotation, document, citation, or source you enter to support a claim. "Position" means a recorded stance on an issue. "Proposed Fix" means a remedy a Citizen proposes for a recorded problem. "Proposed Law" means draft legislative text generated or edited in the Service.

2.4 Money. "Subscription" means a recurring paid entitlement. "Plan" means a level of Subscription. There are no named tiers, and no tier buys a different product. Supporter membership is a single monthly Subscription whose level is the amount you choose: $40, $100, $200, $400 or $1,000 per month, or any other amount you enter. Every Supporter receives the same benefits at $40 per month or more. "Billing Period" means the interval between charges. "Renewal Date" means the date a Subscription automatically renews. "One-Time Purchase" means a non-recurring paid item, including a purchased title on WeTheCitizens.tv. "Rental" means a time-limited right to view a title. "Donation" means a voluntary contribution that is not a purchase of anything (§17.1).

2.5 Integrity. "Inauthentic Account" means an Account that misrepresents who is operating it or on whose behalf, including an Account participating in Coordinated Narrative Control. "Coordinated Narrative Control" means concerted activity by two or more Accounts, or by one Account acting under direction, intended to manufacture the appearance of organic opinion, suppress opinion, or manipulate the Service's reasoning or ranking — whether directed by a government, an intelligence service, a political party, a campaign, a donor, a corporation, a nonprofit, or any other group, powerful or not.

2.6 Accuracy. "Inaccurate Data" means Evidence or User Content that is materially false. "Knowing Falsification" means submitting Inaccurate Data that you know to be false, or that you have deliberately altered, fabricated, or misattributed. It does not include an honest mistake, a typo, a good-faith error, a disputed interpretation, or an opinion (§10.9). "Synthetic Media" means images, audio, or video that are generated or materially altered by software, including AI, so as to depict a real person or real event.

2.7 AI. "AI Output" means text, analysis, drafts, scores, summaries, or media produced by an automated model in the Service, including reasoning checks, evidence-gap assessments, and Proposed Laws. "AI Provider" means a third party whose model or service we use to produce AI Output (§18.7).

2.8 "Board of Directors" means an advisory group. Where the Service, our website, or any of our materials refers to a "board of directors", a "board", or a "board member", that phrase means an advisory group — a group of guides and advisors who give us their view. It does not mean a corporate board of directors. Its members hold no corporate office at ACT 3 AI, Inc., are not directors of it in any legal sense, owe you no fiduciary duty, and have no authority to bind, represent, or speak for ACT 3 AI, Inc. (§29.12). Where we ever do have a statutory board of directors, we will call it that and say so plainly.

3. Scope — One Product, Many Front Doors

In short: Four party front doors, one web app, one contract. And these Terms bind our registered Users — not everyone on the wider federated network (§3.6).

3.1 What these Terms cover. These Terms govern all of the following, which are one product operated by one company: WeTheCitizens.io (the movement application); WeTheCitizens.tv (streaming and pay-per-view); the Local App; and any related application, API, or mobile app we publish.

3.1a Our sibling social network is not covered by these Terms. It is operated by the same company (§1.1), but under its own Terms of Service and its own Privacy Policy — and those documents, not these, govern your use of it. One company, two products, two contracts: the contract that applies is the one published on the property you are actually using. We say it here because two overlapping contracts over one service would be a gap dressed up as thoroughness, and because our Privacy Policy already says the same thing about your data (Privacy Policy §2.1a).

3.2 The four front doors. We operate four audience-specific front doors:

Domain	Audience
WeCitizensR.com	Republicans
WeCitizensD.com	Democrats
WeCitizensL.com	Libertarians
WeCitizensSocialism.com	Democratic Socialists

These are one web application serving a different front page depending on the hostname you arrived on. They are not four products, four companies, or four contracts.

3.3 One contract, no matter which door you came through. The door you enter by does not change your rights, our obligations, the price you pay, the entity you contract with, or the rules that apply to you. These same Terms govern all four, and govern wethecitizens.io and app.wethecitizens.io identically. Nor does the door you came through mean we hold its politics — serving an audience is not adopting its positions (§17.9, §17.13).

3.4 What we are not. We are not a political party, a campaign, a political committee, a PAC or super PAC, a lobbying firm, or a news outlet. Serving an audience is not endorsing its politics, and operating a front door for a political tendency is not adopting it (§17.9). We are also not a nonprofit or a tax-exempt organization (§17.12).

3.5 Feature-specific terms. Some features may carry additional terms presented to you at the point of use. Those supplement these Terms as to that feature and do not replace them (§1.6).

3.6 Who these Terms bind — and who they do not. These Terms are an agreement between us and our registered Users. They do not apply to people who are not registered Users of our Service, and that expressly includes people using other services on the Federated Network who interact with content on our Service, whose content is visible to our Users, or who see content our Users published onward (§19.2).

We say this for three reasons. It is true: we have no contract with those people and never did. It matters to you: much of what you read on the Federated Network, and much of the audience for what you publish there, is people we have no relationship with and no authority over (§5.7, §19.3). And it matters to them: a movement built on consent should not claim power over people who never agreed to anything. Nothing in these Terms gives a person who is not a User any right to enforce them (§29.5), and nothing in them purports to govern that person's conduct.

3.7 Not all of this exists yet, and we would rather tell you which parts.

These Terms describe the Service as it is designed, including parts of it we have not built yet. As of the last-modified date at the bottom of this document:

Payments and Subscriptions (§15) are not live. No payment processor is integrated and we hold no billing data.

WeTheCitizens.tv (§16) does not exist. There is no video service and no viewing data.

Publishing to a federated network (§19) is not live.

The four front doors in §3.2 do not yet operate as separate hostnames.

Self-service export (§20.4) is not built. The right is real and stated; the button is not there yet. Our Privacy Policy §18.8 says the same thing.

The transparency report and the enforcement counters (§13.6, §13.11) are not published yet.

The public change log (§28.4) does not exist yet. Until it does, the last-modified date at the bottom of this document is the only version marker.

The DMCA designated agent (§12.8) is not yet registered with the Copyright Office, so the safe-harbor description at §14.4 states what we are putting in place rather than what is already in place. We will not assert §512 protection we have not established.

Why the terms for them are here anyway. Rules that arrive after a feature does are rules nobody agreed to. Writing them in advance means that on the day payments launch, the auto-renewal disclosure, the cancellation right, and the refund position already exist and were already visible to you — which is the whole point of §15.7. Our Privacy Policy takes the same approach and says the same thing about the same features, in its own sections 2.3, 12 and 16.

What that does and does not mean. A section describing a feature that does not exist imposes no obligation on you and confers no right on you until that feature is available to you — you cannot breach §16 before there is anything to watch. Nothing in these Terms is a promise, commitment, or representation that we will build any feature, on any timetable, or at all (§21.5, §21.6). We will update these Terms and give notice under §28.2 before any of those surfaces launches, and if what we build differs from what is described here, the description changes — not your rights.

4. Eligibility, Accounts, and Identity

In short: One human per account, you are responsible for what happens under yours, and you can use the Local App without an account at all.

4.1 Minimum age. You must be at least 18 years old to create an Account. Under no circumstances may anyone under 13 create an Account or use the Hosted Service. If we learn that an Account belongs to someone below the minimum age, we will close it.

4.2 Accurate registration. Information you give us at registration must be accurate, and you must keep it current.

4.3 One human, one Account. An Account is for a single human being. You may not operate an Account that presents automated activity as a person, share your Account, or sell, rent, or transfer it. Automated agents are permitted only where we expressly provide for them and only if clearly labeled as automated.

4.4 Your credentials, your responsibility. You are responsible for keeping your credentials secure and for all activity under your Account. Tell us promptly if you believe your Account has been compromised.

4.5 Identity is optional for local use. You may use the Local App without an Account. An Account and a verified identity are required for: publishing to the Hosted Service or the Federated Network; participating in a Community; paid Subscriptions and purchases; and features that attribute a Position to a person. We require identity for those because they affect other people, not because we want the data.

4.6 Third-party sign-in. Authentication may be provided through OpenAuthFederated or a third-party identity provider. That provider's own terms and privacy policy also apply to you, and we are not responsible for its acts, omissions, or availability.

4.7 Sanctions and restricted jurisdictions. You may not use the Service if you are located in an embargoed jurisdiction, or if you are a person or entity on a denied-party, sanctions, or restricted-party list. You represent that you are not. Separately, AI features may be unavailable in some countries where our AI Provider does not offer service, even though the rest of the Service is available to you there (§18.14).

4.8 Accounts for organizations. If you create or administer an Account on behalf of a Community or organization, you represent that you have authority to do so and that the individuals whose data you enter have been told and, where required, have consented.

4.9 Use of your credentials is presumed to be you. Each time your credentials are used, we will presume you authorized the activity, and we are not required to verify the source of access. We are not liable for loss, theft, or fraudulent use of your Account. This presumption is rebuttable and it does not apply once you have told us your Account is compromised (§4.4) — but until you tell us, activity under your Account is treated as yours, including posts, purchases, and Subscriptions.

4.10 Age assurance for some features. Where the law of your jurisdiction requires it, we may condition access to particular features or content on completing an age-assurance process, using methods that comply with that law. This is separate from the minimum age in §4.1: §4.1 governs whether you may hold an Account at all, and this section governs whether a particular Account may reach a particular thing (§8.4, §16.9, §26.5).

5. The Local-First Split — What Lives Where

In short: Files on your machine are yours and we cannot see, restore, or hand them over. Anything you publish is a different story, and publishing to the federated network cannot be undone — by us or by you.

This section describes the most important structural fact about the Service. Read it before §20 and before the Privacy Policy.

5.1 Your local files are yours. Positions, values, Evidence, reasoning chains, drafts, and private notes created in the Local App are stored as files on your own computer. We do not receive them, we cannot read them, and we claim no rights in them. They are yours in every sense.

5.2 We cannot back up, restore, or recover local data. We have no backup obligation for anything on your machine, and no technical ability to restore it. If your disk fails, if you delete a file, or if you lose your computer, that data is gone and we cannot help you. Keep your own backups.

5.3 Published data is different, and so is anything the app sends us. Anything you choose to send to the Hosted Service, to a Community, or to the Federated Network is governed by the license in §7 and by our Privacy Policy. The moment you publish, the data crosses from the private side to the shared side, and different rules apply. The Privacy Policy — not this section — is the complete and controlling account of what the Local App transmits to us, including any diagnostic, error-reporting, update-check, or usage information, and of whether that is on by default or requires your opt-in (§20.1). §5.1 describes your Positions, values, Evidence, and notes. It is not a claim that the Local App never communicates with us at all, and you should read the Privacy Policy rather than infer one.

5.4 Publishing to a federated network is not reversible. When you publish to any Federated Network, copies propagate to relays, media servers, and instances operated by other people. Deleting your copy, or ours, does not delete theirs. We will delete our copy and, where a protocol supports it, transmit a deletion request downstream — but a downstream request is not a deletion, and we cannot promise any third party will honor it. Treat publication as permanent. §5.7 says this again, in the words that matter most.

5.5 No promise of forward compatibility. We do not promise that any version of the Local App will keep working with future versions of the Hosted Service, or that file formats will remain unchanged.

5.6 Self-hosted and forked instances. Our source code is open. If you or anyone else runs an independent instance from it, that operator — not us — is responsible for it, and these Terms do not cover it. We make no representation about any instance we do not operate, and a Citizen using someone else's instance is contracting with that operator, not with us.

5.7 Read this before you publish anything.

Because this is a federated service, content you publish may — and probably will — be transferred to other people's servers. Deleting content from our Service does not guarantee that it will be deleted from other servers or other people's computers, because that is outside our control — and outside yours.

We put it that way deliberately, and each part of it is doing work:

"Probably will", not "may". Propagation is the normal case, not the exception. If we wrote "may", you could reasonably read federation as a rare event, and it is not.

"Does not guarantee", not "we are not liable". This is a statement about what is physically possible, not a shifting of risk. We are telling you we cannot, not that we will not.

"Outside our control — and outside yours." The limit is symmetrical. Neither of us can call it back. We would rather you knew that than discovered it.

What we will actually do when you delete: we delete our copy; where the protocol supports it, we transmit the deletion downstream and notify the other services and applications on the network that the content or the Account is gone; and we tell you here, in advance, that some copies may persist anyway. To ask another service to remove something, contact that service directly — we have no authority over it (§19.2, §19.3).

6. License to Use the Service

In short: You get a limited licence to use the hosted Service. The open-source licence on our code is a completely separate document covering a completely different thing — and it does not give you our name. You may fork our code; you may not run it under our brand. You may always use our name to talk about us, including to criticize us.

6.1 Your license. Subject to these Terms, we grant you a limited, revocable, non-exclusive, non-transferable, non-sublicensable license to access and use the Hosted Service for its intended purpose: participating in the movement as a Citizen.

6.2 Restrictions. You may not: reverse engineer, decompile, or disassemble any part of the Hosted Service that is not open source; circumvent rate limits, access controls, paywalls, or technical protection measures; resell or sublicense access; scrape or harvest data except as §12.13 permits; or interfere with the operation or security of the Service.

6.3 The code license is a different document. Our source code is released under an open-source license. That license governs the code and nothing else — it grants you no right to the Hosted Service, to our brand, to other Citizens' content, or to the data on our servers. Equally, these Terms grant you no rights in the code and take none away.

6.4 Our marks. WE THE CITIZENS™ is our trademark. Our logos, our trade dress, and the four front-door brands — WeCitizensR.com, WeCitizensD.com, WeCitizensL.com, and WeCitizensSocialism.com — are also ours, together with the goodwill in all of them (the "Marks").

6.5 Beta and experimental features. Features labeled beta, preview, or experimental are provided with no promises of any kind and may be changed or withdrawn at any time.

6.6 The three-way carve-out. To be explicit, because these three things are commonly and harmfully conflated:

(a) The code is governed solely by its open-source license. Nothing in these Terms restricts any right that license grants you, and the restriction in §6.2 applies only to parts of the Hosted Service that are not open source.
(b) These Terms govern solely the Hosted Service and the properties in §3.1.
(c) Purely local use of the Local App — running it on your own machine with no connection to the Hosted Service — requires no acceptance of these Terms at all and is governed only by (a).

6.7 No implied trademark license — and this sentence sits here, next to the open-source license, on purpose. Nothing in these Terms, nothing in the open-source license on our code, and nothing in your use of the Service grants you any right, license, or permission to use the Marks. We say it here rather than in a distant intellectual-property section because the thing sitting immediately beside it — a permissive open-source license — is exactly what leads a reader to assume the brand came with the code. It did not. Conveying copyright while withholding trademark is the settled convention of open-source licensing; the Apache License 2.0 says so in terms at its §6, and the MIT and BSD licenses reach the same result by saying nothing at all, which is the more surprising route to it.

6.8 What you may not do with the Marks. Without our prior written permission, you may not use the Marks, or any confusingly similar variant, misspelling, translation, or transliteration of them, in any of the following:

a domain name or subdomain;
an account handle, username, or display name;
an application, product, or service name;
a business name, trade name, or assumed name (DBA);
a metatag, keyword advertisement, or search advertisement; or
any other use that suggests we sponsor, endorse, operate, are affiliated with, or have approved you, your instance, your organization, or your content.

6.9 Why this clause carries more weight here than on an ordinary platform. Our code is open source and anyone may lawfully run a fork of it — we want that. But a fork running under our name is, to a citizen or a voter, indistinguishable from us, and everything this movement is worth depends on a person knowing whose position, whose ballot information, and whose proposed legislation they are reading. The Marks are the only legal instrument that separates us from a fork, because the code license deliberately does not. That is why §6.8 enumerates rather than gestures, and it is why a convincing look-alike operated to shape what citizens believe is treated here with the same seriousness as the coordinated narrative-control conduct described in §9 and §14.

6.10 Referring to us, including to criticize us, is expressly permitted. Nothing in §6.7 or §6.8 prevents you from using our name to identify, discuss, review, report on, compare with, or criticize us, our product, or our positions. That is nominative fair use, it is lawful, and we are not pretending otherwise. Two independent reasons, either sufficient on its own: a clause purporting to forbid it would be unenforceable and could take the enforceable parts of §6.8 down with it; and we are a free-speech platform, so a trademark clause that functioned as a gag on our critics would refute the entire premise of this Service.

6.11 Other people's marks. Trademarks, logos, and brands belonging to third parties are their owners' property. Their appearance anywhere in the Service implies no affiliation, sponsorship, or endorsement in either direction. Your own obligations toward other people's marks are in §12.2 and §12.4.

6.12 Reservation of rights. All rights not expressly granted in these Terms are reserved.

6.13 A note on the "™" symbol, because we would rather be accurate than impressive. We use "™" and not "®". "™" asserts that we claim the mark; "®" asserts that a trademark office has registered it. We use the symbol that is true, and we will not upgrade it until a trademark office has actually registered the mark (§12.14).

7. User Content and the License You Grant Us

In short: You own what you write. We get a limited licence to run the Service with it, and that licence ends when you delete it. You also grant a SECOND, separate licence: we use what you publish to train AI — our own models, and models of companies outside ours — and that one is permanent, because a model cannot be untrained. It runs only while your separate consent under the Privacy Policy stands, you can withdraw that consent to stop future training, and anything that never leaves your own machine is never trained on at all. We will never sell your content as a mailing list without asking you first.

7.1 You own your content. You keep all ownership of your User Content. We claim none.

7.2 The license you grant — OPERATING THE SERVICE. For User Content you publish through the Hosted Service, you grant us — and you represent and warrant that you have the right to grant us — a worldwide, non-exclusive, royalty-free, sublicensable, and transferable license to host, store, cache, reproduce, adapt for format and display, publish, transmit, and distribute it for the purpose of operating, securing, and promoting the Service. This license is limited to that purpose. It does not extend to User Content that never leaves your own machine (§5.1) — we have no license to that because we never receive it.

7.2A THE SECOND LICENSE — AI TRAINING, OURS AND OTHER COMPANIES'. This is a separate grant and it works differently from §7.2. Read it.

(a) The grant. You grant us a worldwide, non-exclusive, royalty-free, sublicensable and transferable license to use, reproduce, and create derivative works from your User Content in order to develop, train, fine-tune, evaluate, and improve artificial intelligence and machine learning models — (i) our own models, and (ii) by sublicense, the models of third parties outside our company, for their own AI training.

(b) It is deliberately separate from §7.2, and it is not carried by it. §7.2 is limited to operating, securing and promoting the Service and does not authorize training. This clause does, and only to the extent stated in it.

(c) IT IS PERPETUAL AND IRREVOCABLE — and §7.4 does not end it. A model cannot be untrained. Once your User Content has been used to train a model, that use cannot be reversed, and this license survives deletion of the content, deletion of your Account, and termination of these Terms. §7.4's duration rule expressly does not apply to this clause.

(d) WHAT YOU CAN ACTUALLY DO ABOUT IT, and it is real. This license is exercised only so far as your separate consent under the Privacy Policy allows (§7.2C). You may withdraw that consent at any time, which stops all future training — while, per (c), not undoing training already performed. And the surest control is architectural: User Content that never leaves your own machine is never trained on by anyone, because we never receive it (§5.1). This clause reaches only what you publish or submit to us.

(e) Scope. The categories of User Content covered are those named in the Privacy Policy's training section, and no others.

7.2B Sublicensees are bound. Any third party we sublicense under §7.2A(a)(ii) is contractually limited to AI training, may not re-share your User Content, and may not use it to identify or contact you.

7.2C THE LICENSE IS NOT THE CONSENT — and we are not pretending otherwise. §7.2A is our contractual right. It is not the legal basis for processing your political opinions, and accepting these Terms does not give that basis. Political opinions are special category data; processing them for training requires separate explicit consent, obtained outside these Terms, unbundled from acceptance and separately withdrawable, as set out in the Privacy Policy. Where that consent is absent or withdrawn, §7.2A is not exercised regardless of what these Terms would otherwise permit. If these Terms and the Privacy Policy ever appear to conflict on this point, the Privacy Policy controls (§20.1).

7.3 Onward publication at your direction. The license in §7.2 includes distributing your published User Content to the destinations you select: the Hosted Service and the Federated Network.

7.4 Duration. The license in §7.2 ends when you delete the User Content or your Account, except as §7.10 provides for backups, and except that: (a) it survives for copies already distributed to the Federated Network, which we cannot recall (§5.4, §5.7); and (b) it survives for content we must retain for legal compliance, abuse prevention, or an open moderation or legal matter.

This clause does not apply to §7.2A. The AI-training license is perpetual and irrevocable by its own terms (§7.2A(c)), because a model cannot be untrained. Deleting your content or your Account stops future training; it does not and cannot reverse training already performed. This is the same kind of one-way door as publication to the Federated Network (§19.3), and we would rather state it here than let §7.4 imply a reversibility that does not exist.

7.5 Moral rights. To the extent permitted by law, you waive moral rights in your published User Content as against ordinary operation of the Service. Where such rights cannot be waived, you agree not to assert them against us for acts within the scope of §7.2.

7.6 Feedback. Suggestions, feature requests, and bug reports you send us are ours to use freely and without obligation or compensation. This includes reports you send us under §13.12.

7.7 No compensation. We do not pay you for User Content unless a separate written agreement says otherwise.

7.8 Public means public. Anything you publish may be seen, copied, quoted, screenshotted, archived, indexed by search engines, and republished by anyone. Do not publish anything you are not prepared to have permanently public.

7.9 The warranty is part of the grant. §7.2 folds your rights warranty into the licence sentence itself, and that is deliberate. You cannot grant us a licence you do not hold. If you publish content you had no right to publish, you have not merely breached §12.1 — you have purported to grant a licence that was never yours to give, and the consequences of that are yours (§24.3, §24.4).

7.10 Your licence ends automatically when you delete — with one honest exception for backups.

When you delete User Content, the licence you granted us in §7.2 to keep hosting, showing, and transmitting that content ends automatically. You do not have to ask us, and we do not have to agree.

The exception, stated because it is true rather than because it helps us: deleted content may remain in our backups for a period. For that content, and for that purpose only, the licence does not terminate automatically. If we restore the Service from a backup that contains it, we will once again be hosting and transmitting it. To reduce the chance of that: we restore from the most recent working backup, and if we do restore from a backup we will announce it through the Service, so that you can check whether anything you deleted has come back and delete it again.

Backup retention periods are in the Privacy Policy (§20.1, §20.5). Copies already on the Federated Network are a different problem entirely and are governed by §5.7 — no deletion of yours or ours reaches those.

7.11 We will never sell your content without asking you first — and here is exactly what that does and does not cover.

What it covers, absolutely: we will not sell, rent, trade, or license your User Content as a mailing list, a contact list, a voter file, or a marketing audience, and we will not sell it to any third party for any purpose, without getting your permission first. This is a limit on §7.2 and it is not narrowed by anything else in these Terms.

What it does not cover, stated plainly rather than left to inference: the AI-training sublicense at §7.2A(a)(ii), under which your User Content is provided to parties outside our company for their AI training. That is authorized by §7.2A and by your separate consent under the Privacy Policy — it is not authorized by this clause, and this clause does not forbid it. Whether that arrangement constitutes a "sale" or "share" in the technical sense used by the California Consumer Privacy Act is addressed in the Privacy Policy §8.1, which controls on that question. Your permission for it is asked separately, and you may refuse or withdraw it without losing the rest of the Service.

This clause is also separate from — and does not authorize — the anti-scraping and AI-training prohibition we impose on other people at §12.13.

8. Acceptable Use

In short: We are serious about free expression and we wrote that into this contract. We also keep full discretion to moderate — and both halves are true at once.

8.1 Free expression is a commitment we put in the contract. We hold free expression as a core commitment of this movement, and we state it here, in the contract you are accepting, rather than only in our marketing. We do not remove content because we disagree with it, and we do not remove content because it is unpopular, uncomfortable, or critical of powerful people or institutions. Criticizing officials and institutions is the purpose of this platform.

8.2 Sexual content involving minors. Absolutely prohibited, in any form, real or synthetic, including drawings, renderings, and AI-generated material. Zero tolerance: immediate termination, preservation of the content and associated data, and reporting to the National Center for Missing & Exploited Children as required by 18 U.S.C. § 2258A (§14.5).

8.3 Non-consensual intimate imagery. Prohibited: intimate imagery shared without the subject's consent, so-called revenge pornography, and sexual Synthetic Media depicting a real person.

8.4 Adult content. Pornographic and obscene material is not permitted on the Service.

8.5 Gore and graphic violence. Prohibited: real-world injury, death, or mutilation posted for shock value. Documentary and newsworthy material about real events is permitted, and may be placed behind a warning or age gate (§16.9, §13.1, §4.10).

8.6 Violence and violent extremism. Prohibited: true threats of violence, incitement to violence, terrorist and violent-extremist content, and material support for a designated terrorist organization.

8.7 Doxxing. Prohibited: publishing a real person's home address, personal phone number, government identification number, financial account number, or comparable private information. Publishing an official's public office contact information is not doxxing.

8.8 Harassment. Prohibited: targeted abuse, stalking, sustained unwanted contact, and coordinated brigading of an individual. Vigorous criticism of a public figure's public conduct is not harassment.

8.9 Technical attacks. Prohibited: malware, exploits, phishing, credential theft, and any attack on the security or availability of the Service or on another Citizen.

8.10 Spam and manipulation. Prohibited: bulk unsolicited messaging, engagement farming, vote and ranking manipulation, and automated scraping outside §12.13.

8.11 Illegal goods, fraud, and scams. Prohibited.

8.12 Impersonation. Prohibited: impersonating a real person, official, agency, or organization. Clearly labeled parody, satire, and commentary are permitted.

8.13 Unlabeled synthetic media. Synthetic Media depicting a real person or real event must be labeled as such (§10.3).

8.14 Election interference. Prohibited: false voting instructions, false polling places, dates, or eligibility rules, and voter intimidation. Arguing about elections is protected; lying to people about how to vote is not.

8.15 We may moderate; we are not obliged to monitor. We have no general obligation to monitor User Content and we do not pre-screen it. Reviewing some content does not create a duty as to all content and does not make us responsible for what we did not review. §13.15 describes work we choose to do; choosing to do it does not convert this section into a duty of care.

8.16 Why we nonetheless keep discretion — and how the two fit together. Precisely because we permit a broad range of speech, we must retain the discretion to act. A platform that hosts hard speech attracts people who will use it for things no one should have to host, and prudence sometimes requires us to err on the side of caution. We therefore reserve full discretion over our content moderation, exercised at our sole judgment, and we may remove content, limit its reach, label it, or suspend or terminate an Account. We narrow that discretion in one respect that matters: termination is for violation of these Terms, for legal risk, or under §9 — not for holding a view we dislike. The constraints in §13 (strikes, notice, appeal windows, published counters) are how we make that commitment checkable rather than merely stated.

8.17 We act on reports. We do review reports we receive and act on them in good faith. We do not commit to a particular outcome, a particular timeline, or to acting on every report, but we do not operate a reporting mechanism that goes nowhere.

9. Inauthentic Accounts and Coordinated Narrative Control

In short: We deliberately look for governments, intelligence services, and political money running coordinated narrative control here, and you agree we may judge that at our discretion. That authorization is the whole point of this section.

9.1 We look for it deliberately. We will intentionally and actively consider whether a government, an intelligence service, a political party, a campaign, political money, a corporation, a nonprofit, or any other group — powerful or not — is conducting Coordinated Narrative Control on the Service. This is a stated purpose of our integrity work, not an incidental one.

9.2 It is treated as inauthenticity. Coordinated Narrative Control is treated as operating an Inauthentic Account, and is a violation of these Terms by every Account participating in it.

9.3 We categorize it at our sole discretion, and you agree that we may. Determining what constitutes Coordinated Narrative Control requires judgment about intent, coordination, and provenance that cannot be reduced to a mechanical rule. You expressly agree that we may make that determination at our sole discretion, and that acting on it is authorized by this contract and is not a breach of it.

9.4 Signals we may consider. Without limitation: coordinated timing or phrasing; shared infrastructure or provenance; undisclosed direction, funding, or employment; behavior inconsistent with a single human operator; artificial amplification or suppression; and manipulation of evidence or reasoning. This list is illustrative and is not a limit — a closed list would be a roadmap for evasion.

9.5 Consequences. Removal, reach limitation, labeling, suspension, or termination. Where an Account is terminated under this section, paid time is forfeited (§13.7, §15.10).

9.6 No disclosure of methods, no advance notice. We are not obliged to disclose our detection methods or to warn an Account before acting under this section. Explaining exactly how detection works to the people evading it would defeat the purpose.

9.7 No evasion. You may not evade enforcement, including by creating a new Account after termination, or by operating additional Accounts to circumvent a limit.

10. Data Accuracy and Integrity

In short: Be honest with evidence. Knowingly faking it costs $400,000 per incident. Honest mistakes cost nothing and you can always correct them.

This Service exists to reason from evidence. Falsified evidence does not merely mislead one reader — it poisons every conclusion downstream of it and every Citizen who relied on that chain. Accuracy is therefore a contractual obligation here, not an aspiration.

10.1 Your warranty of accuracy. You represent and warrant that every fact, figure, quotation, document, citation, and piece of Evidence you enter is accurate to the best of your knowledge, and that you have not knowingly falsified, fabricated, doctored, or misattributed any of it.

10.2 Faithful transcription and quotation. When you transcribe or quote a source — a speech, a video, a hearing, a document, a broadcast — your transcription must be faithful. You may not: edit selectively so as to reverse or materially change the meaning; fabricate a quotation; attribute words to a person who did not say them; or alter a document and present it as original.

10.3 Synthetic media must be labeled. If you post AI-generated or materially altered imagery, audio, or video depicting a real person or a real event, you must label it. An unlabeled synthetic depiction of a real person is treated as Inaccurate Data under this section and may also violate §8.13 and §11.

10.4 Sourcing. Factual claims should carry a source. Opinion should be identifiable as opinion. This is how a Citizen reading your reasoning can check it, and it is the ordinary standard of the Service.

10.5 Why this section is stricter than you may expect. The Service builds chains of reasoning across many Citizens' contributions. A single deliberately falsified item can invalidate conclusions far from where it was entered, require re-verification of everything that touched it, and expose both us and other Citizens to claims by the person it was used against.

10.6 LIQUIDATED DAMAGES — $400,000 PER INCIDENT.

⚠️ PLEASE READ THIS PROVISION CAREFULLY. IT REQUIRES YOU TO PAY A SUBSTANTIAL SUM IN SPECIFIED CIRCUMSTANCES.

(a) Trigger — knowing conduct only. If you engage in Knowing Falsification as defined in §2.6 — that is, you submit Evidence or User Content that you know to be false, or that you have deliberately altered, fabricated, or misattributed — you agree to pay us US $400,000 for each such incident.

(b) This provision does not apply to mistakes. It does not apply to an honest error, a typographical error, a misremembered detail, a source that later turns out to be wrong, a disputed interpretation, an opinion, or a good-faith factual claim that proves incorrect. It applies only to deliberate falsification. See §10.9.

(c) Why this is a reasonable estimate and not a penalty. The parties acknowledge that our actual damages from deliberate falsification are real but genuinely impracticable to determine in advance, because they include: the cost of identifying and re-verifying every reasoning chain, Position, and Proposed Fix that relied on the falsified item; the cost of notifying affected Citizens and correcting the record; investigation and remediation cost; the loss of trust in a platform whose entire value proposition is that its evidence is honest, which is a diffuse and compounding harm rather than a countable one; and our exposure to claims by third parties defamed or damaged by the fabrication. The parties agree that the sum in (a) is a reasonable pre-estimate of those damages, agreed at the time of contracting, and is NOT a penalty and is not intended to punish.

(d) Reduction rather than voiding. If a court, arbitrator, or mediator determines that the amount in (a) is unenforceable as written, the parties intend that it be reduced to the maximum amount that is enforceable, and that this provision be given effect as so reduced, rather than struck out entirely. See §29.2.

(e) Our other remedies are preserved. Nothing in this provision limits our right to seek actual damages, injunctive relief, or any other remedy, in the alternative or in addition (§10.8).

(f) This provision is ours, not yours. §10.6 runs only in our favor and imposes no corresponding liability on us. Our own liability to you is capped by §22 and is not affected by this section.

(g) Our cap does not cap this. For the avoidance of doubt, and as §22.14 states from the other direction, the limitation of liability in §22.3 limits what WE may owe YOU. It does not limit what you may owe us under this section. A cap on our liability is not a cap on yours, and reading it as one would empty this section of content.

10.7 Our ordinary remedies. Short of §10.6, we may remove or annotate content we reasonably believe is falsified, re-verify affected reasoning, mark or restrict the Account, suspend or terminate it, and recover our costs. We may do any of that at our discretion and we do not owe you a hearing before we do it — though the appeal route in §13.4 and §13.10 remains open to you afterwards.

10.8 Cumulative remedies. Our remedies under this section are cumulative and not exclusive.

10.9 Mistakes are expressly excluded. To repeat, because it matters: an honest mistake is not a breach of this section. We would rather Citizens enter their best understanding and correct it than stay silent for fear of liability.

10.10 Correcting the record. You may correct anything you got wrong, at any time, without penalty, through the correction tools in the Service, or by telling us at the address in §30 (§13.12). Prompt voluntary correction is evidence of good faith and weighs against any finding of Knowing Falsification.

10.11 Context for §10.6. (Research note, not an operative clause.) Across the forty-six comparable consumer contracts read for this document, exactly one carries a liquidated-damages figure of its own. It is materially smaller than ours and is tied to a mechanically countable breach rather than to a contested state of mind. §10.6 is drafted the way it is — knowing conduct only, an express damages recital, an express "not a penalty" statement, a reduce-rather-than-void fallback, and an explicit carve-out for honest error — precisely because that is thin precedent to stand on. This is flagged for outside counsel (Appendix D.1).

11. Defamation and Statements About Real People

In short: Criticize the powerful all you like — that is what this is for. Do not state false facts about real people. And you agree not to sue us over what someone else said.

11.1 You will not defame anyone. You may not publish false statements of fact about a real person or a real organization. Specifically prohibited: fabricated quotations; doctored documents or media presented as authentic; and knowingly false accusations of criminal conduct.

11.2 The line is fact, not tone. Opinion, criticism, satire, parody, and commentary are protected and welcome here, however harsh. What is prohibited is a false statement of fact. "This official's policy is corrupt and incompetent" is opinion. "This official took a $50,000 bribe on 4 March" is a factual assertion, and you had better be able to support it.

11.3 Public figures and officials. Scrutinizing officials, institutions, corporations, and public figures is the purpose of this platform and we encourage it without reservation. To do it safely and credibly: label opinion as opinion, cite sources for factual claims (§10.4), and distinguish what you know from what you infer.

11.4 You will not sue us for defamation. You agree not to assert any claim for defamation, libel, slander, false light, injurious falsehood, or business disparagement against us based on: (a) User Content published by another Citizen; (b) our moderation decisions, including removal, labeling, reach limitation, or refusal to remove; (c) labels, fact-flags, warnings, or annotations we apply; or (d) AI Output. This does not waive any claim you may have against the Citizen who actually made the statement, and it does not waive claims that cannot lawfully be waived (§23.5).

11.5 You indemnify us for defamation claims arising from your content. Your indemnity under §24 expressly includes claims by third parties for defamation and related torts arising from your User Content, including our reasonable attorneys' fees. This obligation survives termination of your Account — defamation claims routinely arrive years after the content was posted and long after an Account is gone (§24.7, §27.7).

11.6 If someone says you got it wrong about them. Any person who believes a false statement of fact has been published about them may report it through the channel in §30 — including a person who has no Account with us at all. We will review it. We do not promise removal and we do not promise refusal — we exercise the discretion in §8.16 and §13, taking into account the evidence supplied by both sides.

11.7 Correction and retraction. Where a factual claim is shown to be false, we may require correction or annotation, and we may correct or annotate it ourselves. The Citizen who posted it should correct it under §10.10.

11.8 Outside the United States. Defamation law in the United Kingdom, the European Union, Australia, and elsewhere is materially less favorable to publishers and platforms than US law, and in some of those jurisdictions truth alone is not a complete answer. If you publish about people in those jurisdictions, you bear that risk (§26.3–§26.6).

11.9 Naming a real person in a Position, Proposed Fix, or campaign. Where you name an identifiable private individual as a subject of a Position, Proposed Fix, or campaign — as distinct from commenting on a public figure's public conduct — we may require you to demonstrate that person's consent or a legitimate public-interest basis. You will have 7 days from our request to do so, and until you do we may withhold the associated feature, payout, publication, or distribution. This is deliberately a deadline attached to a withheld benefit rather than one more warranty, because warranties change no one's behavior.

12. Intellectual Property, Infringement, and DMCA

In short: Only upload what you have the right to upload. Here is how to send a copyright notice, and here is our repeat-infringer policy.

12.1 You must hold the rights. You represent and warrant that you own, or hold every license, right, consent, and permission necessary for, all User Content you upload, and that it infringes no one's rights. See also §7.9 — you cannot grant us a licence you do not hold.

12.2 No infringement. You may not upload content that infringes copyright, trademark, patent, trade secret, publicity, privacy, or moral rights.

12.3 Third-party material. Do not upload someone else's article, photograph, music, film clip, or footage unless you have the right to do so.

12.4 Marks. Do not use another party's trademarks or logos in a way that suggests affiliation, sponsorship, or endorsement.

12.5 Right of publicity. Do not use a real person's name, likeness, or voice — including a synthetic reproduction of any of them — to imply that they endorse a Position, a party, a candidate, or us. The same rule applies to AI Output: that a model mentioned someone does not mean they endorse anything, and you may not present it as though it did (§18.15).

12.6 Fair use is your call and your risk. Whether a particular use is fair use, fair dealing, or otherwise permitted is your judgment and your risk. We do not make that assessment for you and we do not indemnify it. Before you upload a government document, a hearing recording, a news report, or any other third-party source, it is your responsibility to check that you may lawfully post it here.

12.7 DMCA notification. If you believe content on the Hosted Service infringes your copyright, send a written notice to our designated agent (§12.8) containing: (a) a physical or electronic signature of the owner or authorized agent; (b) identification of the copyrighted work claimed to be infringed; (c) identification of the material claimed to be infringing and information reasonably sufficient to locate it; (d) your contact information; (e) a statement that you have a good-faith belief the use is not authorized by the owner, its agent, or the law; and (f) a statement, under penalty of perjury, that the information is accurate and that you are authorized to act.

12.8 Designated agent. Send copyright notices to ContactUs@ACT3ai.com (subject line: "DMCA"), or by post to ACT 3 AI, Inc., Attn: Copyright Agent, 16430 NE 50th Street, Redmond, WA 98052, United States of America. This agent is not yet registered with the United States Copyright Office (§3.7), and until that registration is complete we do not assert the §512 safe harbor (§14.4). We will act on valid notices regardless — the registration governs what we may claim, not what we will do.

12.9 Counter-notification. If your content was removed and you believe it was removed in error or by misidentification, you may send a counter-notification containing the elements required by 17 U.S.C. § 512(g)(3), including your consent to the jurisdiction of the federal district court for your address (or, if outside the United States, any district in which we may be found). If we receive a valid counter-notification, we may restore the material in 10–14 business days unless the original complainant notifies us that it has filed an action seeking to restrain the activity.

12.10 Repeat infringers. We will terminate the Accounts of repeat infringers in appropriate circumstances. This process is not yet in operation (§3.7, §14.4), and we will not describe it as running before it is. It is a condition of the §512 safe harbor rather than a courtesy, which is why we state its status plainly instead of asserting a protection we have not established (§14.4).

12.11 Misrepresentation. Under 17 U.S.C. § 512(f), a person who knowingly materially misrepresents that material is infringing, or that it was removed by mistake, may be liable for damages. We take abusive notices seriously in both directions.

12.12 Trademark complaints, and why they go somewhere else. Trademark complaints go through the trademark row of the §30 table, not to the DMCA agent, and are handled separately from the DMCA process. The reason is that the DMCA is a copyright statute and its safe harbor does not cover trademark at all — a trademark grievance sent to a copyright agent is a notice nobody is obliged to act on, and a complainant who is told the right address is better served than one left waiting. Tell us the mark, the registration if there is one, where it appears on the Service, and why the use confuses. Our own marks are in §6.4.

12.13 Scraping and AI training. You may not access the Service by automated means, or use User Content obtained from it, for any commercial purpose — and "commercial purpose" expressly includes training, fine-tuning, or supplying data to any artificial intelligence system, including large language models and generative AI — except (a) as our published API terms permit, (b) as our robots.txt permits, or (c) with our prior written consent. Nothing in this section restricts a Citizen from exporting their own data (§20.4), and nothing here overrides rights granted by the open-source license on our code (§6.6). This prohibition binds other people; it is not a licence to us to do the same thing, which §7.11 forecloses.

12.14 We hold ourselves to §12.2 as well. §12.2 forbids you to infringe a trademark; this section states what we owe you in return on the same subject. We will not describe any mark of ours as registered unless a trademark office has actually registered it, which is why §6.13 uses "™" and not "®", and we will not assert a right in a mark we do not hold. A platform that polices other people's trademark conduct while overstating its own is not one you should take at its word.

13. Moderation, Enforcement, and Appeals

In short: What we can do, the strike system and its expiry, real appeal deadlines that run against us too, how to tell us we got something wrong, and how to complain about us.

13.1 What we may do. We may remove content, restrict its reach, label or annotate it, place it behind a warning or age gate, demonetize it, suspend an Account, or terminate an Account.

13.2 Discretion, and its limits. We act at our sole discretion, with or without prior notice. We are not generally obliged to explain a decision — except where §26.4 (EU Digital Services Act) or other applicable law requires a statement of reasons, in which case we will provide one.

13.3 Reporting abuse. Any Citizen may report content or conduct that breaks §8. We review reports and act on them in good faith (§8.17). We do not commit to a specific outcome or timeline. Reports about a mistake rather than about abuse go to §13.12.

13.4 Appeals are real, and are not a right to reinstatement. You may appeal an enforcement decision. We will consider it in good faith. An appeal is not a guarantee of reinstatement, and our decision on appeal is final as between us, subject to §25 and to §26.4.

13.5 Preservation and legal process. We may preserve and disclose content and Account information where required by law, valid legal process, or where we believe in good faith that disclosure is necessary to prevent imminent physical harm.

13.6 Government and law-enforcement requests. We require valid legal process. We will resist requests we consider overbroad or improper. Where we are lawfully permitted to do so, we will notify the affected Citizen. We publish a periodic transparency report. Note what we cannot produce: we do not hold your local files and cannot hand them over, because we never received them (§5.1).

13.7 No refund on termination for breach. If we terminate an Account for breach of these Terms or under §9, paid time is forfeited and is not refunded (§15.10).

13.8 Enforcement is not a legal conclusion. Removing content is not an admission that it was unlawful, and leaving content up is not an endorsement of it or a statement that it is lawful.

13.9 Strikes, and strikes expire. Enforcement actions against an Account are recorded as strikes. A strike expires 90 days after it is issued. An Account that accumulates three live strikes within any 180-day period may be terminated. A strike notice will name the specific rule and identify the specific content at issue — a notice that says only "you violated our policies" is not good enough and we will not send one.

13.10 Appeal windows, as actual numbers, running in both directions. You have 15 days from a strike notice to appeal. We will answer an appeal within 30 days of receiving it. If we do not answer within 30 days, the appeal is treated as unresolved and §25.11 is satisfied as to that matter. We state our own deadline as well as yours on purpose: a clock that runs only against the user is the first thing anyone challenging this section would point at, and it would deserve to be pointed at.

13.11 We publish the counters. We publish running totals of enforcement actions taken, appeals received, and appeals granted. A platform that asks Citizens to trust its discretion should be willing to be counted.

13.12 Tell us when we get something wrong — the corrections route. We work to make content on the Service accurate, and we will sometimes fail (§18.12, §21.9). If you find a mistake — in AI Output, in a summary, in a label we applied, in a citation, or anywhere else — please tell us at the corrections address in §30. We ask you to do this because it genuinely helps. Tell us what is wrong, where you found it, and what you believe is correct. What we do with a report is described in the Accuracy and Corrections Policy (§1.5, §1.9).

What this route is and is not. It is how we learn about errors and, where we can, fix them. It is not a promise to correct, to respond, or to reach any particular result, it creates no duty of care and no liability on our part, and it does not change §18.12, §21 or §22. Reporting a mistake does not make us the publisher of what we did not write (§14.2) and does not turn our discretion into an obligation (§8.15).

13.13 The internal complaint process. If you have a complaint about us — about an enforcement decision, a charge, a feature, a person on our staff, or anything else — this is how it works, and §25.11 requires you to use it before starting any other proceeding:

(a) How to file, and what to put in it. Send it to the complaints address in §30, or use the complaint form in the Service. So that we can actually deal with it, tell us the things listed at §25.15 — who you are, how to reach you, that you are invoking this process, what happened and when, what you want, and the evidence you have.
(b) Our deadline. We will acknowledge within 7 days and give you a substantive written answer within 30 days of receiving a complete complaint.
(c) When it ends. The process ends on the earlier of: our written answer; your written withdrawal; or 30 days after you filed. It never runs longer than 30 days, and if we miss that deadline the condition precedent in §25.11 is satisfied automatically and you are free to proceed.
(d) It costs you nothing. We do not charge for it and we do not require you to use a lawyer.
(e) It does not take away your other routes. Filing a complaint does not waive any claim, does not extend or shorten any deadline against you, and does not stop you using the carve-outs in §25.13. All limitation periods are tolled while it runs (§25.14). And, to be explicit: this process supplements, and does not replace, your right to go to court or to complain to a regulator (§25.17).

13.14 What an appeal or complaint has to contain to be workable. An appeal under §13.4 or a complaint under §13.13 must be well founded and must include enough information or documentation for us to conduct a meaningful review. An appeal that gives no supporting detail, or that arrives after the window in §13.10, may not be eligible for reconsideration. This is not a trap: §25.15 lists exactly what to include, and if you send that, it is enough. And it cuts both ways — if we miss our own deadline in §13.10 or §13.13(b), you do not lose anything; the condition simply falls away (§25.12).

13.15 Work we do before anyone reports anything. We use a combination of automated detection and human review to find prohibited material, and we prioritize the fastest removal of child sexual exploitation material, terrorist and violent-extremist content, and other priority illegal content. That work may include matching against databases of known illegal material and automated detection of content that violates §8. We also act on reports from Citizens, from trusted flaggers, from law enforcement, and from civil-society organizations.

Why this paragraph is here, and what it does not do. We describe this work because §14.2's protection depends on our moderation being in good faith, and because a platform that says nothing about proactive safety work invites the assumption that it does none. But describing it does not create an obligation to do it, does not make it complete or effective, and does not create a duty of care to you or to anyone else (§8.15, §21.3). Automated detection is imperfect in both directions, and we do not promise otherwise.

14. Our Legal Role — Safe Harbors

In short: You wrote it, not us. We say what Section 230 does and does not cover, and we do not lean on it alone.

14.1 We are an interactive computer service. We provide an interactive computer service within the meaning of 47 U.S.C. § 230. You are the information content provider of your own User Content. We do not create it and do not adopt it.

14.2 Moderating does not make us the publisher. Under 47 U.S.C. § 230(c)(2), our good-faith decisions to restrict access to material do not make us the publisher or speaker of material we did not create. §8.17 records that we act on reports in good faith, and §13.15 records the proactive work that good faith consists of; that is the predicate this protection rests on. Neither our review of a corrections report under §13.12, nor our answer to a complaint under §13.13, nor any proactive detection under §13.15, is an adoption of the underlying content.

14.3 What Section 230 does not cover. Section 230 does not apply to federal criminal law, to intellectual property claims, to the FOSTA sex-trafficking exceptions, or to certain state-law claims. These Terms do not rely on Section 230 alone, and the obligations you accept under §10, §11, §12 and §24 stand independently of it.

14.4 DMCA safe harbor. We do not currently assert the §512 safe harbor, because two of its conditions are not yet met: the designated agent at §12.8 is not registered with the Copyright Office, and the repeat-infringer process at §12.10 is not yet in operation (§3.7). When both are in place we will operate under 17 U.S.C. §512 — maintaining a designated agent (§12.8), responding to valid notices (§12.7), honoring counter-notifications (§12.9), and terminating repeat infringers (§12.10). We would rather claim nothing than claim a protection we have not established.

14.5 Mandatory reporting overrides expectations of privacy. Where we are required to report apparent child sexual abuse material under 18 U.S.C. § 2258A, we will preserve and report it, and no expectation of privacy, confidentiality, or notice applies to that content.

14.6 No editorial control asserted, no endorsement. We do not assert editorial control over User Content and we do not endorse it. Content on the Service is the Citizen's speech, not ours.

14.7 We are not a common carrier. We are not a common carrier, a public utility, a public forum, or a state actor, and nothing in these Terms — including the commitment in §8.1 — makes us one or creates any must-carry obligation. §8.1 is a contractual commitment we chose to make; it is not a surrender of the discretion reserved in §8.16.

15. Payment, Credit Cards, and Subscriptions

In short: Subscriptions auto-renew until you cancel; cancelling is as easy as signing up was; we never pre-check a recurring box; deleting your account does not cancel billing; and we do not refund on request — except where your law says we must.

⚠️ THIS SECTION CONTAINS AUTOMATIC RENEWAL TERMS. YOUR PAYMENT METHOD WILL BE CHARGED AUTOMATICALLY, ON A RECURRING BASIS, UNTIL YOU CANCEL. YOU MAY CANCEL AT ANY TIME. SEE §15.3 AND §15.4. FEES ARE GENERALLY NON-REFUNDABLE — SEE §15.10 AND §15.24.

15.1 Authorization to charge. You authorize us and our payment processor to charge your payment method, on a recurring basis, for the Plan price, applicable taxes, and any fees, until you cancel.

15.2 You may hold more than one subscription, and cancelling one does not cancel the others. You may hold several Subscriptions at the same time — more than one Plan, seats in more than one Community, and a WeTheCitizens.tv purchase or Rental alongside a movement Plan. Each Subscription is billed on its own cycle and must be cancelled separately. Cancelling one Subscription does not cancel any other, and deleting your Account does not cancel any of them (§15.20). If you want to stop paying entirely, cancel each Subscription.

15.3 Automatic renewal. Each Subscription renews automatically at the end of each Billing Period for a further period of the same length, at the then-current price, unless you cancel before the Renewal Date. At purchase we will show you, clearly and conspicuously: the renewal period, the renewal price, and the deadline for cancelling, and we will send you an acknowledgement you can keep.

15.4 Cancel any time, in no more steps than it took to sign up. You may cancel any Subscription at any time, from inside the Service, in no more steps than signing up required. We will not require a phone call, an email, a chat session, or a retention conversation. Cancellation stops the next charge.

15.5 What happens after you cancel. Your access continues to the end of the period you have already paid for, and then stops. You may ask us to end access immediately instead; that does not create a refund entitlement (§15.10).

15.6 We never pre-check a recurring box. Recurring billing is opt-in by an affirmative act. We do not pre-select, pre-check, or default any charge to recurring, and we do not convert a one-time payment into a recurring one. This is a commitment, not a description of current UI, and it applies to Donations as well (§17.2).

15.7 Auto-renewal law compliance. We intend this section to satisfy the federal Restore Online Shoppers' Confidence Act (ROSCA) and state automatic-renewal laws, including California's Automatic Renewal Law: clear and conspicuous disclosure before the charge; affirmative consent to the recurring charge; an acknowledgement you can retain; advance notice before a renewal at a changed price or following a free trial; and a simple, immediate cancellation mechanism.

15.8 Free trials and promotional rates. If we offer a trial or an introductory rate, we will tell you before you accept: what happens when it ends, the amount that will be charged, and the date of the first charge. We will notify you in advance of the first charge at the standard rate. We do not silently convert a trial into a paid Subscription.

15.9 Price changes. We may change prices. An existing Subscription keeps its price until its next Renewal Date. We will give you at least 30 days' notice before a renewal at a new price, with an opportunity to cancel first.

15.10 Refunds. Except where law requires otherwise, fees are non-refundable and partial Billing Periods are not refunded. If we choose to give a goodwill refund, that is discretionary, sets no precedent, and creates no entitlement. No refund is due where an Account is terminated for breach or under §9 (§13.7).

15.11 Failed payments. If a charge is declined we may retry it over a limited period and may suspend the Subscription. A retry after a declined charge does not move your Renewal Date — your billing cycle stays where it was (§15.23). After a grace period we may downgrade or terminate the Subscription. You remain responsible for amounts already owed.

15.12 Taxes. Prices exclude sales, use, value-added, goods-and-services, and similar taxes unless stated. You are responsible for them.

15.13 Chargebacks. Filing a chargeback instead of using the cancellation process (§15.4) or asking us for a refund (§15.10) is a breach of these Terms. We may suspend the Account and recover the disputed amount together with the processor's dispute fees.

15.14 Payment processor, and exactly what card data we hold. Payments are handled by a third-party processor. We do not receive or store your full card number. What we do receive and hold is a payment token from the processor, the last four digits, the card brand, the expiry date, and the billing address — we say that specifically rather than claiming we "do not store payment information", which would not be quite true. The Privacy Policy is the controlling statement of it (§20.1). The processor's own terms and privacy policy apply to you in addition to these Terms, it is a separate controller for its own purposes, and we are not liable for its acts, omissions, outages, or security.

15.15 Currency, descriptor, receipts. Prices are stated in the currency shown at checkout. We will tell you the descriptor that appears on your statement, and we will provide a receipt for every charge.

15.16 Past-due amounts. We may recover reasonable costs of collection on past-due amounts, and interest where permitted by law.

15.17 Plan changes mid-period. On an upgrade, the change takes effect immediately and we charge a prorated amount for the remainder of the Billing Period. On a downgrade, the change takes effect at the next Renewal Date and no refund is due for the current period.

15.18 Seats paid for by an organization. Where an organization pays for seats, that organization is responsible for the fees, and its administrator may add or remove seats. Losing a sponsored seat does not entitle you to a refund from us.

15.19 Each paid item is its own contract. Each time you order a paid item — a Subscription, a One-Time Purchase, a Rental, or a seat — you enter into a separate contract with us for that particular item. Because they are separate contracts, they have separate Renewal Dates, separate cancellations, and separate consequences. If you have purchased access to several things, you may have several renewal dates.

15.20 Deleting your Account does not cancel your Subscriptions. Cancel each Subscription first, then delete your Account. If you delete an Account with a live Subscription, billing may continue under §15.19. §27.9 says you may leave at any time without telling us; this is the one thing you should do before you go.

15.21 Mandatory local law prevails. Some states and countries give you mandatory cancellation and refund rights. Nothing in this section overrides those rights, and where they conflict with anything above, they win. That includes anything in §15.10 and §15.24.

15.22 Everything about billing is in this contract. We deliberately keep our billing terms here, in the main agreement you accepted, rather than in a separate payments document you would never open. If we ever add a supplementary payments document, it will supplement this section and will be presented to you before you pay (§1.6, §3.5).

15.23 No silent movement of your renewal date. We will not move your Renewal Date without telling you. Retries (§15.11), plan changes (§15.17), and promotional periods (§15.8) either preserve your existing cycle or are disclosed to you before they take effect.

15.24 WE DO NOT REFUND MONEY MERELY BECAUSE YOU ASK.

We do not agree to give money back on request. A request for a refund is not, by itself, a reason for one, and we are under no obligation to grant it. Asking us, being dissatisfied, changing your mind, forgetting you had a Subscription, or not having used what you paid for are none of them grounds for a refund.

What this does not do. It does not affect a refund the law of your jurisdiction requires us to give (§15.21, §26.6), a statutory cooling-off or cancellation right, a refund we owe under §16.8 for a title you paid for and could not watch, or the pro-rata refunds we commit to at §27.6, §27.8 and §28.2. Those still stand. We say this because several jurisdictions give a consumer a refund right that a contract cannot switch off, and a clause pretending otherwise weakens the whole section instead of strengthening it.

If you think a charge was a mistake, use §13.13 before you use your card issuer (§15.13).

16. Streaming, Pay-Per-View, and Rentals (WeTheCitizens.tv)

In short: "Buy" means a licence to watch, not a copy you own — and a title can go away when our rights to it end. Rentals have a clock. We say all of this up front.

16.1 Free and paid tiers. Some titles are free to watch. Others require a Subscription, a Rental, or a One-Time Purchase.

16.2 A purchase is a licence, not ownership. When you "buy" a title, you are buying a limited, personal, non-transferable licence to view it through the Service. You are not buying a copy and you do not own the title. You cannot resell, lend, or bequeath it.

16.3 Rental windows. A Rental gives you a period to start watching and a shorter period to finish once you have started. Both periods are shown before you pay. When the window closes, access ends. See §16.12 for how those limits are expressed.

16.4 Devices, streams, and geography. We may set limits on registered devices, concurrent streams, and household use, and titles may be unavailable in some countries because of the rights we hold. Limits in force are shown in the Service.

16.5 Titles can go away, including ones you paid for. Our right to offer a title comes from a licence granted to us by a rightsholder. When that licence ends, the title may become unavailable — including to Citizens who purchased it. We will give notice where we reasonably can. Where a purchased title becomes unavailable within a reasonable period after purchase and you have not substantially viewed it, contact us under §30 and we will make it right.

16.6 No copying or circumvention. You may not download (except through a download feature we provide), record, redistribute, publicly perform, or circumvent any technical protection measure.

16.7 Quality and availability are not guaranteed. Streaming quality depends on your connection, device, and location. We do not guarantee any resolution, bitrate, or uninterrupted playback.

16.8 Refunds for titles you cannot watch. If a title you paid for is unwatchable because of a fault on our side, we will refund or replace it. This is the one refund we commit to outright, and §15.24 does not switch it off. Dissatisfaction with the content, a slow connection on your side, or an expired Rental window you did not use are not grounds for a refund.

16.9 Warnings, ratings, and age gates. We may apply content warnings, ratings, and age gates, and may require age verification for particular titles (§4.10).

16.10 We will tell you the truth about what "buy" means. To be explicit, because this is where streaming services most often mislead people: a purchase lasts only for as long as we retain the rights (§16.5); a viewing window may be shortened if our underlying rights change, and where that happens we will not charge you again for the same title; and purchases end when your Account ends (§27.4). If you find that unacceptable, do not purchase — rent or subscribe instead.

16.11 We have no obligation to keep a title available, and we are not liable if it goes.

We have no responsibility to continue making any title available to you through the Service, and we will not be liable to you if a title — including a title you purchased — becomes unavailable.

Note the order of those two things, because it is deliberate. First: we do not owe you a duty to keep a title up. Only second: we are not liable if it goes. A clause that disclaimed only the liability would leave open the argument that we owed you the duty all along and simply refused to pay for breaking it.

And here is the mechanism, named rather than hidden: a title becomes unavailable because the licence a rightsholder granted us ends, or because we lose the right to make it available in your country. That is the ordinary life of licensed video, it will happen, and §16.5's make-it-right commitment is what we offer against it.

16.12 Rental limits, stated as one sentence. A Rental is viewable on a single device at a time, must be started within the start window shown to you before you pay, and must be finished within the completion window shown to you before you pay, which begins when you first press play. Both windows and the device limit are disclosed at the point of purchase (§16.3, §16.4). Once the completion window has begun it runs whether or not you are watching, and an unused window is not refundable (§16.8).

17. Donations, Fundraising, and Campaign Finance

In short: A donation is a gift, not a purchase, and it is not tax-deductible because we are not a nonprofit. We host political speech; we are not a pipe for political money.

17.1 A donation is not a subscription. A Donation buys you nothing. A Subscription is a purchase of access. We will never describe one as the other, will never present a Donation as a purchase, and will never present a purchase as a Donation.

17.2 Recurring donations get the same protections as subscriptions. A recurring Donation requires the same affirmative opt-in (§15.6), the same conspicuous disclosure (§15.3), the same one-step cancellation (§15.4), and the same advance notice (§15.8, §15.9) as a paid Subscription. We do not pre-check recurring donation boxes. We consider the practice — well documented among political fundraising platforms, and the subject of regulator and legislative attention — to be indefensible, and we do not engage in it.

17.3 Refunds and tax treatment. See §17.11 (a Donation is not refundable on demand) and §17.12 (nothing is tax-deductible). There is no political action committee and no political committee of any kind anywhere in our structure (§17.14).

17.4 We do not endorse candidates. We do not endorse or oppose any candidate, party, or campaign. Content on the Service is the Citizen's, not ours (§14.6).

17.5 We are not a conduit for political money. You may not use the Service to make, solicit, route, or process a contribution to a candidate, political committee, party committee, or PAC, unless and until we expressly build that capability and comply with the applicable Federal Election Commission rules for it. Our payment rails are for Subscriptions, purchases, and Donations to us — not for political contributions.

17.6 No foreign national contributions or foreign-directed electioneering. You may not use the Service to make or direct any contribution or electioneering communication prohibited by 52 U.S.C. § 30121 or comparable law.

17.7 No coordination that would make us an in-kind contributor. You may not use the Service to conduct coordinated campaign activity that would cause us to make a prohibited or unreported in-kind contribution. Candidates, campaign committees, party committees, and PACs may not monetize through the Service or use it as a fundraising channel.

17.8 Your own compliance is yours. You are responsible for your own compliance with campaign finance, lobbying disclosure, and election law in every jurisdiction that applies to you. We do not advise you on it (§18.5).

17.9 Non-partisan by construction. The four front doors (§3.2) are audiences we serve, not positions we hold. Operating WeCitizensR.com does not make us Republican and operating WeCitizensSocialism.com does not make us socialist. We built four doors because citizens of every tendency deserve the same tools.

17.10 Host the politics; do not be the conduit for the money. §17.5 and §17.7 exist to draw a bright line: political speech on the Service is broadly permitted and protected; political money through the Service is not. The two are separate questions and we answer them differently.

17.11 A donation is a gift to support the work, and it is not refundable on demand.

When you donate, you are giving us money to support this work. You are not buying anything, we are not holding it for you, and it does not become refundable because you later change your mind or ask for it back. We do not agree to return a Donation on request (§15.24).

As with §15.24, this does not switch off a refund the law requires, and it does not apply to a charge that was a genuine error — a duplicate, an amount you did not authorize, or a recurring Donation you never opted into. Tell us at §30 and we will fix an error like that, because a Donation nobody knowingly made is not a Donation.

17.12 WE ARE NOT A NONPROFIT AND NOTHING YOU PAY US IS TAX-DEDUCTIBLE.

We make no claim of nonprofit, charitable, or tax-exempt status. ACT 3 AI, Inc. is not represented to be a 501(c)(3), a 501(c)(4), or any other tax-exempt organization, and we do not hold ourselves out as one.

No payment to us — no Subscription, no purchase, no Rental, and no Donation — is a tax-deductible charitable contribution. We will not issue a charitable-contribution acknowledgement, and you should not treat any payment to us as deductible. If your tax position matters to you, ask your own adviser; we do not give tax advice (§18.5).

This statement is repeated at the point of donation, because that is where a donor will actually read it.

17.13 Featuring something is not endorsing it. The appearance of a Position, Proposed Fix, campaign, Community, or any other User Content on the Service — including on a front door listed in §3.2, in a ranked or recommended list, in a newsletter, or anywhere we have chosen to surface it — does not constitute an endorsement, sponsorship, or affiliation of or with it by us.

We disclaim featuring, and not merely hosting, on purpose. Disclaiming what we passively store is easy and answers the easy case. The case that actually arises is content we surfaced — and a Citizen's Position appearing on WeCitizensR.com is not our position, our party's position, or a statement that we agree with a word of it (§3.3, §17.9, §14.6).

17.14 If a political committee ever exists in our structure, we will say so and say where. There is no political action committee and no political committee of any kind anywhere in our structure (§17.3). If one ever does, we will name that entity, name the property it operates, and say plainly which entity you are dealing with on which hostname — rather than letting one contract quietly cover two entities with different legal characters. Until that statement appears here, you are contracting with ACT 3 AI, Inc. alone, and no payment you make to us is a political contribution (§17.5).

18. AI Features and AI Output

In short: The AI can be false, incomplete, misleading, or out of date. Check it. It is your speech once you publish it, we take no liability for it, and there is a real address for telling us it got something wrong.

18.1 What the AI does. The Service uses automated models to check chains of reasoning against first principles, identify gaps in Evidence, compare Positions to a reference social contract, and draft Proposed Laws and summaries.

18.2 AI Output is software output, not fact and not advice. AI Output is not a statement of fact, not professional advice, and not our opinion. It is generated text.

18.3 The AI can be wrong, and will sometimes be confidently wrong. Factual assertions in AI Output should not be relied upon without independently checking their accuracy, as they may be false, incomplete, misleading, or not reflective of recent events or information. AI Output may also be biased or fabricated, and may be delivered with unwarranted confidence. You must verify AI Output before relying on it for anything that matters. For a product about current politics, "not reflective of recent events" is the failure mode to watch hardest.

18.4 Proposed Laws are drafts for discussion. A Proposed Law generated in the Service is a discussion draft. It is not a legal instrument, not legislative drafting advice, and not a substitute for a lawyer. §18.17 sets out, in full, what that means and why.

18.5 No professional advice, anywhere in the Service. Nothing in the Service — from us, from the AI, or from another Citizen — is legal, tax, accounting, financial, investment, medical, psychological, election-law, or safety advice.

18.6 Training — WE DO TRAIN, and the details are in the Privacy Policy. Your User Content is used to train AI models — ours, and those of parties outside our company — under the license at §7.2A and subject to your separate consent under §7.2C. The categories covered, the recipients, the consents, and your controls are governed by the Privacy Policy, which controls on every question of what we do with personal data (§20.1). Where your Positions or other special-category information are used for training, that requires your separate, explicit, opt-in consent, which is not given by accepting these Terms (§7.2C).

Do not read §18.7 as a no-training promise. Where a third-party model provider's own terms commit that provider not to train on what we send it, that is a commitment by them. It is not a commitment by us, and we do train. Both facts are true at once.

18.6A Training is irreversible. A model cannot be untrained. Withdrawing consent stops future training and does not remove your content from a model already trained on it, ours or a third party's. We will never describe a training-data deletion request as a deletion — the same rule we apply to the Federated Network at §19.3. §18.18 explains why we do not make a blanket promise here.

18.7 Third-party AI providers. Some AI features are provided by third-party AI Providers, whose terms, usage policies, retention practices, and availability apply to those features. We are not liable for their outages, changes, or errors. §18.14 sets out what that means for you.

18.8 Prohibited uses of the AI. You may not use AI features to generate: defamatory factual claims about real people (§11); Synthetic Media of real people that you then publish unlabeled (§8.13, §10.3); or fabricated Evidence (§10). You must also comply with the usage policy of any AI Provider whose model your request reaches (§18.14).

18.9 Ownership of AI Output. As between you and us, you may use AI Output you generate, subject to these Terms. §18.16 states precisely what that does and does not mean.

18.10 Credits, quotas, and limits. AI features may be subject to credits, quotas, or rate limits, which we may change. When you exhaust them, those features stop until the next period or until you purchase more. Unused credits do not carry over unless we say so.

18.11 We are not the publisher or speaker of AI Output. Where you prompt, direct, edit, adopt, or publish AI Output, you are the speaker of it. For all purposes, including §11 and §14, we are not to be treated as the publisher or speaker of AI Output generated at your direction, AI Output you publish is treated as your User Content, and you assume sole responsibility for it — including responsibility for verifying it before you publish (§18.3).

18.12 WE TRY TO GET IT RIGHT. WE USE AI. AI MAKES MISTAKES. WE ARE NOT LIABLE FOR THE RESULT.

All four of those sentences are true at once, and we put them in one place so that none of them can be read without the others.

(a) We make an effort. We work to make content presented through the Service accurate, and we build tools — reasoning checks, evidence-gap tracking, sourcing prompts — whose whole purpose is accuracy. We are not indifferent to whether the Service tells you the truth.

(b) That effort is not a warranty, and you must not rely on it. Making an effort is not a promise of a result. Nothing in (a) is a representation, warranty, or guarantee that any content, Evidence, summary, label, or AI Output is accurate, complete, current, or fit for any purpose, and nothing in (a) creates a duty of care to you (§21).

(c) AI is involved, and AI gets things wrong. Automated models produce errors, omissions, distortions, and outright fabrications, sometimes stated with complete confidence (§18.3). That is a known and unavoidable property of the technology, not an aberration.

(d) WE TAKE NO LIABILITY FOR THE RESULT. To the fullest extent permitted by law, we are not liable for any loss, damage, decision, or consequence arising from an inaccuracy in any content or AI Output, whether or not we made an effort to prevent it and whether or not we were told it might happen. Our total liability, if any, remains capped by §22.

(e) Please tell us. There is a real address for it. We ask you to report mistakes at the corrections address in §30 (§13.12). That request is an invitation, not an assumption of responsibility: reporting a mistake to us, and our reviewing it, creates no obligation to correct, no timeline, and no liability (§13.12).

18.13 Why §18.3 is written the way it is — and why it does not depend on us. The warning in §18.3 is not a stylistic choice and it is not us being cautious.

The commercial terms on which AI capability of this kind is supplied require the customer to notify its users that factual assertions in model output should not be relied upon without independently checking their accuracy, because they may be false, incomplete, misleading, or not reflective of recent events or information. Where we operate on those terms, that notification is an obligation we owe our provider, not a courtesy we extend to you.

We are stating it that way on purpose. Which provider we use, and on which commercial surface, is a fact about our own arrangements that we have not yet confirmed, and the Privacy Policy says the same. We are not going to assert a vendor contract we have not confirmed in order to make a disclaimer sound more official. So read §18.3 as standing on three independent legs, any one of which is enough: it is required of us where those terms apply; the enumeration is a better description of how these models actually fail than anything we would have written; and it is our own position regardless (§18.12, §21.9). You are entitled to the warning either way.

18.14 The AI Provider's rules apply to you too.

When you use an AI feature, your request reaches an AI Provider. You are, in that provider's terms, one of its users — and anything you do that breaks its rules is our breach of contract with it. So:

You agree to comply with the usage policy of any AI Provider we use, as we make it available or link it, in addition to §8 and §18.8. Where the two differ, the stricter applies to the AI feature.

AI features may be unavailable in some countries even where the rest of the Service is available, because AI Providers support a narrower list of regions than we do (§4.7). That is a limitation of the feature, not a suspension of your Account.

We may change AI Provider, or stop offering an AI feature, at any time (§6.5, §21.6).

If you break an AI Provider's rules, we may suspend your access to AI features immediately and separately from any other enforcement (§13.1), and you indemnify us for the consequences (§24.3).

18.15 A mention is not an endorsement. A reference to a third party in AI Output does not mean that person or organization endorses, is working with, or is associated with us, with you, or with anything the output says. Do not present AI Output as though a named person had said, approved, or endorsed it — that is both a violation of §12.5 and, if it is a false statement of fact, of §11.1. This is the AI form of the same rule §17.13 states for content we feature.

18.16 Ownership of AI Output, stated precisely.

As between you and us, whatever rights exist in AI Output you generate are yours, and we assign to you whatever right, title, and interest (if any) we have in it, subject to your compliance with these Terms.

Read "(if any)" carefully — it is the honest part of the sentence. We are not representing that copyright or any other right subsists in AI Output at all. In several jurisdictions, material generated by a machine without sufficient human authorship attracts no copyright protection, and we make no representation that yours does. You may therefore find that AI Output you generated is not exclusively yours as against anyone else, and that other Citizens generate something similar.

We promise you no indemnity for AI Output, and here is the reason, stated rather than hidden. The protection our AI Provider offers us against third-party intellectual-property claims does not extend to claims arising from the inputs a user supplies, from the combination of the model's output with other content, or from modifications made to the output — which, between them, describe nearly every realistic claim this product could generate. We therefore have nothing to pass on to you, and we do not pretend to. Your obligation under §24 runs the other way and is unaffected.

18.17 Anything the Service says about law — including a Proposed Law.

There is no assurance that any statement in the Service touching on legal matters is true, correct, or precise. Law varies from place to place and it changes over time, sometimes quickly. Even a statement that was accurate may be accurate only in the jurisdiction of whoever entered it, and the law may since have been amended, superseded, or overturned. Those three failure modes — wrong jurisdiction, out of date, superseded — are exactly the ones an automated model is worst at noticing (§18.3).

Legal information in the Service is general in nature and cannot substitute for the advice of a licensed professional who can apply the law to your actual circumstances. If you need that, contact a local bar association, law society, or comparable body and ask for a referral. We would rather point you somewhere useful than merely tell you not to rely on us.

Neither we, nor our officers, employees, contractors, advisors, contributors, or the developers of the Service, take any responsibility for the results or consequences of any attempt to use or adopt information — or misinformation — presented through the Service.

Nothing in the Service is an attempt to offer or render a legal opinion or otherwise to engage in the practice of law, and no attorney-client relationship arises between you and us in any circumstance.

18.18 What we will and will not say about training — and why the difference matters.

Where an AI Provider's commercial terms commit that it will not train its models on the content we send it, we will say so, name the provider, and name the commitment — in the Privacy Policy, which controls (§18.6, §20.1).

What we will not do is make a blanket promise. "Your inputs are never used to train AI" is the kind of sentence that is true of one provider, on one commercial tier, on the day it is written — and becomes false the moment any part of the product runs through a different provider or a different tier. A privacy claim that quietly stops being true is a misrepresentation, and it is a materially worse outcome for you than a narrower promise we can actually keep.

AND THE HALF THAT MATTERS MOST: A PROVIDER'S PROMISE IS NOT OUR PROMISE. Everything above is about what an AI Provider does with content we send it. It says nothing about what we do. We train AI on your published User Content — our own models, and, by sublicense, models of companies outside ours (§7.2A, §18.6). Those two facts are true at the same time, and we put them in the same clause so that neither can be read without the other. If you read "the provider does not train on it" and concluded that nobody does, that conclusion is wrong.

So: the specific, current, provider-named position is in the Privacy Policy. If it is not written there, do not assume it.

19. Third Parties, Federation, and Integrations

In short: Other people's servers are other people's servers. We cannot delete what has already spread to them, and neither can you.

19.1 Third-party services and links. Links, integrations, and third-party content are not ours, are not endorsed by us, and are governed by their own terms.

19.2 The federated network is run by other people. Relays, media servers, and other instances are operated by independent parties under their own rules and their own privacy practices. We do not control them and are not responsible for them. Their users are not our Users and these Terms do not bind them (§3.6).

19.3 We cannot delete what has propagated. Once your content reaches machines we do not operate, we cannot delete it, correct it, or control it. We will delete our copy and, where the protocol supports it, transmit a deletion request and notify other services on the network — but a deletion request is not a deletion (§5.4, §5.7). To have content removed from another service, contact that service directly.

19.4 Publishing onward. Publishing to any federated destination you select is a choice you make, and it is effectively irreversible. Consider it permanent before you press publish.

19.5 APIs. Use of our APIs is subject to our published API terms and to rate limits, and we may revoke API access at any time.

19.6 App stores. Where a mobile app is distributed through an app store, that store's rules also apply, and the store is not a party to these Terms and has no obligation to support the app.

20. Privacy and Data Handling

In short: The Privacy Policy is the controlling document on anything to do with your data. This section is a signpost to it, not a restatement of it.

20.1 The Privacy Policy controls. Our handling of personal data is governed by the Privacy Policy, which is incorporated by §1.5 and controls over this section on any question of what we do with personal data. This section is a signpost, not a restatement. That includes any question about what the Local App transmits to us — diagnostics, error reports, update checks, usage data — which the Privacy Policy answers and this document does not (§5.3).

20.2 What the Hosted Service collects. Account information, subscription and billing records (§15.14), published User Content, social graph, technical and log data, and support correspondence. Detail is in the Privacy Policy.

20.3 Local data is not ours. We do not collect the Positions, values, Evidence, and notes on your machine, cannot disclose them, and cannot produce them under legal process (§5.1, §13.6). Read that together with §5.3: it is a statement about your content, not a claim that the Local App never communicates with us at all.

20.4 Export. You will be able to export your Account data and your published User Content at any time, in a portable format. The self-service export endpoint is not built yet (§3.7); until it is, ask at the privacy address in §30 and we will produce the export by hand. The Privacy Policy states the same thing at its §18.8, and where the two ever differ on what exists today, the Privacy Policy is the accurate one.

20.5 Deletion. Deleting your Account removes it and your published User Content from the Hosted Service, except for: content already propagated to the Federated Network (§19.3, §5.7); records we must keep for tax, accounting, or legal compliance — and those retention obligations override a deletion request, which the Privacy Policy explains; moderation and abuse-prevention records; and backups until they rotate in the ordinary course, subject to §7.10. Periods are in the Privacy Policy.

20.6 Security. We use reasonable technical and organizational measures to protect data on the Hosted Service. No system is perfectly secure and we do not promise that yours will never be compromised.

20.7 Breach notification. If a breach affects your personal data, we will notify you and the relevant authorities as and when the law requires.

20.8 Cookies and analytics. Our use of cookies and analytics on the hosted properties is described in the Privacy Policy and the Cookie Policy.

21. Disclaimers of Warranty

In short: The Service is provided as is. We warrant nothing about accuracy, and you agree you are not relying on us. Read this one and §22 together.

⚠️ THE SERVICE IS PROVIDED "AS IS". PLEASE READ THIS SECTION AND §22 CAREFULLY — TOGETHER THEY LIMIT WHAT YOU CAN RECOVER FROM US.

21.1 As is, as available. THE SERVICE IS PROVIDED "AS IS" AND "AS AVAILABLE", WITH ALL FAULTS AND WITHOUT WARRANTY OF ANY KIND.

21.2 Disclaimer. TO THE FULLEST EXTENT PERMITTED BY LAW, WE DISCLAIM ALL WARRANTIES, EXPRESS, IMPLIED, AND STATUTORY, INCLUDING MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NON-INFRINGEMENT, ACCURACY, COMPLETENESS, QUIET ENJOYMENT, AND ANY WARRANTY ARISING FROM COURSE OF DEALING OR USAGE OF TRADE.

21.3 No warranty about content, evidence, or AI. WE MAKE NO WARRANTY ABOUT ANY USER CONTENT, ANY EVIDENCE, ANY POSITION, ANY PROPOSED LAW, ANY AI OUTPUT, OR ANY CLAIM MADE BY ANY OTHER CITIZEN. We do not verify User Content and do not adopt it. Nor do we warrant that the proactive work described at §13.15 finds anything, finds everything, or is accurate in either direction.

21.4 No reliance. YOU AGREE THAT YOU ARE NOT RELYING ON US, ON THE SERVICE, ON ANY AI OUTPUT, OR ON ANY OTHER CITIZEN'S CONTENT IN MAKING ANY DECISION — including how to vote, what to file in court, what to publish, what to buy or sell, what medical or safety action to take, or what to do about your own legal exposure. Those decisions are yours, and their consequences are yours.

21.5 No outcome is promised. WE DO NOT PROMISE ANY POLITICAL, LEGISLATIVE, ELECTORAL, OR SOCIAL OUTCOME. We do not promise that your Position, Proposed Fix, or Proposed Law will be read, adopted, acted on, or noticed by any official, institution, or other Citizen.

21.6 No uptime, preservation, or compatibility promise. We do not promise that the Service will be available, uninterrupted, timely, secure, or error-free; that data will be preserved; or that anything will remain backward compatible.

21.7 Savings clause. SOME JURISDICTIONS DO NOT ALLOW THE EXCLUSION OF CERTAIN WARRANTIES. TO THAT EXTENT, THE EXCLUSIONS ABOVE DO NOT APPLY TO YOU, and you may have rights that these Terms cannot remove (§26.6).

21.8 Conspicuousness. This section and §22 are presented in emphasized text deliberately. They are material terms and we intend them to be conspicuous, as the law requires.

21.9 The accuracy disclaimer, stated on its own. WE WORK TO MAKE CONTENT ACCURATE. WE USE AI. AI MAKES MISTAKES. WE DISCLAIM ALL LIABILITY FOR THE RESULT, AND YOU AGREE NOT TO RELY ON THE ACCURACY OF ANYTHING IN THE SERVICE WITHOUT CHECKING IT YOURSELF. Our effort toward accuracy (§18.12(a)) is not a warranty of accuracy, does not create a duty of care, and does not narrow §21.1–§21.4. If you find a mistake, please tell us at the corrections address in §30 — we want to know, and telling us costs you nothing and obliges us to nothing (§13.12, §18.12(e)).

22. Limitation of Liability

In short: Our liability to you is capped at what you have actually paid us. If you have paid us nothing, that is zero, and we say so out loud rather than letting you find out.

22.1 No indirect damages. TO THE FULLEST EXTENT PERMITTED BY LAW, WE WILL NOT BE LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, EXEMPLARY, OR PUNITIVE DAMAGES.

22.2 No lost profits or data. WE WILL NOT BE LIABLE FOR LOST PROFITS, LOST REVENUE, LOST DATA, LOST GOODWILL, LOST OPPORTUNITY, OR THE COST OF SUBSTITUTE SERVICES, however caused and on any theory of liability, even if we were advised of the possibility.

22.3 THE CAP — WHAT YOU PAID US, AND NOTHING MORE.

EXCEPT FOR THE MATTERS LISTED IN §22.12, WHICH ARE NOT SUBJECT TO ANY CAP: OUR TOTAL AGGREGATE LIABILITY TO YOU, FOR ALL CLAIMS ARISING OUT OF OR RELATING TO THESE TERMS OR THE SERVICE, IS LIMITED TO THE TOTAL AMOUNT YOU ACTUALLY PAID US — BY CREDIT CARD OR BY ANY OTHER PAYMENT METHOD — IN THE TWELVE (12) MONTHS IMMEDIATELY BEFORE THE EVENT GIVING RISE TO THE CLAIM.

WE WILL NEVER BE OBLIGATED TO YOU FOR MORE THAN THAT AMOUNT, ON ANY THEORY, AND YOU MAKE NO CLAIM TO ANYTHING MORE THAN THAT.

IF YOU HAVE PAID US NOTHING, THAT AMOUNT IS ZERO. Read §22.9, §22.10 and §22.12 before you assume what that means.

22.4 The cap applies across all theories. The cap in §22.3 applies in the aggregate across all claims and all theories — contract, tort, negligence, strict liability, statute, or otherwise — and is not multiplied by the number of claims, by the number of Subscriptions you hold, or by the number of properties in §3.1 you used.

22.5 Essential basis of the bargain. The disclaimers in §21 and the limits in §22 are an essential basis of the bargain between us, and we have entered into these Terms in reliance on them. The Service is offered on these terms, and most of it is offered free of charge, precisely because liability is limited. Without these limits we could not offer it.

22.6 What we do not exclude. Nothing in these Terms excludes or limits our liability for death or personal injury caused by our negligence, for fraud or fraudulent misrepresentation, or for any other liability that cannot lawfully be excluded or limited. We state this plainly rather than attempting an exclusion we know would not hold. §22.12 sets the same list out as a named, itemised carve-out ahead of the cap.

22.7 Time limit. See §25.7.

22.8 Savings clause. SOME JURISDICTIONS DO NOT ALLOW CERTAIN LIMITATIONS OF LIABILITY. TO THAT EXTENT, THE LIMITS ABOVE DO NOT APPLY TO YOU (§26.6), and to that extent our liability is limited to the greatest extent permitted by the law that applies to you rather than being unlimited (§22.13).

22.9 What §22.3 means for a Citizen who paid us nothing — stated plainly. Most Citizens use We the Citizens for free, locally, and never pay us anything. For those Citizens the cap in §22.3 computes to zero, which means that, to the fullest extent the law allows, we accept no monetary liability to them at all. We are stating that in the open rather than leaving it to be discovered: it is the deal, and a free service is the consideration for it. What it does not touch is anything in §22.6 or §22.12, anything §21.7, §22.8 or §26.6 preserves, or your right to stop using the Service, to complain to a regulator, or to bring a claim we are not permitted to limit.

22.10 If a court will not enforce a zero cap. A cap that resolves to zero is the kind of clause a court examines closely, and some courts will not enforce one against a consumer. If a court or other tribunal holds that §22.3 is unenforceable as applied to you because it produces no recovery, the parties intend §22.3 to be reduced to the minimum amount that is enforceable in that jurisdiction, and to be given effect as so reduced — not struck out, so that the remainder of §22 continues to apply (§29.2). We would rather have a cap a court will enforce than a number a court will delete.

22.11 Where §22.3 sits among comparable services. (Context, not an operative clause.) A cap measured by what the user actually paid, with no fixed floor — so that it resolves to nothing for a user who paid nothing — is the drafting used by at least one long-established non-commercial political advocacy organization, whose terms cap liability at "the amount paid, if any". It is not an outlier confined to platforms nobody wishes to be compared with. It is also not evidence that any such cap has survived a consumer challenge, and §22.10 exists precisely because it may not. Whether to restore a fixed floor is a business decision rather than a drafting one; see §22.9.

22.12 WHAT IS NOT CAPPED AT ALL — read this before you read the cap.

The cap in §22.3 and the exclusions in §22.1–§22.2 do not apply to any of the following, and we list them here, by name, ahead of the cap rather than in a savings clause at the end:

Death or personal injury caused by our negligence.

Our fraud or fraudulent misrepresentation.

Our gross negligence or wilful misconduct.

Our criminal misconduct.

Any statutory right of yours that cannot be contractually disclaimed under the law that applies to you (§26.6).

We put this list before the cap deliberately. A limitation of liability is judged in large part on whether a reader could reasonably have understood it, and a carve-out buried three paragraphs after the number is worth less than one stated before it.

22.13 If a limit fails, it is reduced, not deleted. Where any exclusion or limitation in this section is held unenforceable in a jurisdiction, it applies to the greatest extent that jurisdiction permits, and the remainder of this section remains in full force. This is the same principle as §22.10 and §29.2, applied section-wide.

22.14 This section limits what WE owe YOU. It does not limit what you owe us.

The cap in §22.3 is a limit on our liability to you. It is not a limit on your liability to us, and it does not cap, reduce, or offset: your indemnity under §24; the liquidated damages under §10.6; amounts you owe us for Subscriptions, purchases, taxes, chargebacks, or collection costs under §15; or our right to recover our costs under §10.7.

We say this expressly because a cap that is silent on direction invites the argument that it runs both ways — and if it did, it would empty §10.6 and §24 of any content. It does not. §10.6(g) says the same thing from the other side.

23. Release and Covenant Not to Sue

In short: You release us over what other Citizens do and what you do with what you found here. Politicians and influencers agree to the same thing, said out loud. And we admit what this cannot legally reach.

23.1 Release. You release us from claims, demands, and damages of every kind arising out of or connected with: (a) your own User Content; (b) User Content published by another Citizen; (c) a dispute between you and another Citizen, a Community, or an Area; and (d) anything you did or failed to do in reliance on information you found in the Service (§21.4).

23.2 Covenant not to sue — and its honest limits.

You agree not to bring or assert a claim against us arising from the causes listed in §23.1. This covenant is deliberately narrow: it reaches those causes and nothing else.

This covenant does NOT waive, and does not purport to waive: any claim that cannot lawfully be waived; any right you have under a consumer-protection statute that preserves it; any claim for death or personal injury caused by our negligence; any claim for fraud or fraudulent misrepresentation (§22.6, §22.12); or your right to complain to a regulator or supervisory authority.

We write it this narrowly on purpose. A release that overreaches gets struck down as a whole, taking the enforceable part with it, and we would rather have a clause that works than one that merely sounds strong.

23.3 We are not a party to disputes between Citizens. We are not a party to, and have no duty to investigate, mediate, or resolve, any dispute between you and another Citizen, a Community, a government, a party, a campaign, an employer, a news outlet, or any other third party. This is a different question from §25.10, which is about mediating a dispute between you and us; the two do not conflict and both apply.

23.4 Waiver of unknown claims. Where you are a California resident, you waive California Civil Code § 1542, which reads: "A general release does not extend to claims that the creditor or releasing party does not know or suspect to exist in his or her favor at the time of executing the release and that, if known by him or her, would have materially affected his or her settlement with the debtor or released party." You waive comparable provisions of any other jurisdiction, to the extent such waiver is lawful there.

23.5 Carve-outs preserved. §23.2's carve-outs apply to this entire section, including §23.6 and §23.7.

23.6 PUBLIC FIGURES, OFFICIALS, CANDIDATES, AND INFLUENCERS — THE SAME COVENANT, SAID OUT LOUD.

This provision applies to you, in your personal capacity as an Account holder, if you are an elected or appointed public official; a declared or prospective candidate for public office; an officer, employee, or authorized agent of a political party, campaign, campaign committee, PAC, or government body; or a public figure, commentator, creator, or influencer with a substantial public audience, whether or not you are paid for it. §23.7 explains why the words "in your personal capacity" are there and what happens when they do not apply.

(a) Complain to us — we want you to. You may object to anything on the Service: content about you, a moderation decision, a label or fact-flag, a ranking, or AI Output. Use §11.6, §13.12, or §13.13. We will read it, and being a public figure does not put you at the back of the queue or the front of it.

(b) But you agree not to sue us over it. You agree not to bring or assert any claim against us arising from: your own User Content; User Content published by another Citizen about you or about anyone else; our moderation decisions, including a decision to remove, to label, to limit reach, or to refuse to remove; the labels, warnings, fact-flags, and annotations we apply; or AI Output. This covers, without limitation, claims for defamation, libel, slander, false light, injurious falsehood, business disparagement, tortious interference, and unfair competition.

(c) What it does not reach, and we say so rather than pretend. This covenant is subject to §23.2, §23.5 and §23.7 in full, and additionally does not reach: any claim that cannot lawfully be waived in your jurisdiction; any claim against the Citizen who actually made the statement, which is untouched and is where such a claim belongs; any right you have under an election, campaign-finance, or public-records law; your right to complain to a regulator; or any right the EU Digital Services Act or comparable law gives you to contest a moderation decision (§25.13, §26.4).

(d) Why this is here. This platform exists so that citizens can scrutinize the powerful. A platform that can be sued by the people it is built to scrutinize cannot do that job. We ask for this covenant only from people who chose public life or a public audience, and we ask for nothing here that we do not also ask of every other Citizen in §23.1 and §23.2 — we simply say it explicitly, so that nobody signs it by accident.

23.7 §23.6 binds a person, not an office — and we will not pretend otherwise.

§23.6 applies to you as an individual Account holder. It does not, and does not purport to, bind a government entity, a public body, or an official acting in an official capacity, in any case where the law of the relevant jurisdiction does not permit that entity or that official to give such a waiver.

Why we say this instead of drafting it flat. Many public bodies are barred by statute from agreeing to indemnify, to arbitrate, to accept a foreign forum, or to release claims, and an individual official cannot bind the office by clicking a checkbox. A covenant drafted as though they could would not become enforceable against them — it would simply be evidence that we overreached, and the whole of §23 is written on the principle (§23.2) that an overreaching release takes the enforceable part down with it.

What that means in practice. Where you hold an Account personally, §23.6 binds you personally. Where a government entity uses the Service officially, §29.14 governs and displaces whatever it must. Where the two overlap — an official with a personal account who also posts officially — §23.6 reaches the personal capacity and §29.14 reaches the official one, and no part of this contract asks a public body to give up something it cannot give up.

23.8 We will not imply that a government endorses us. We will not use a government seal, emblem, insignia, official title, or name in a way that suggests any government, agency, or official endorses, sponsors, or is affiliated with us or the Service. This is the mirror of the rule we impose on you at §12.4 and §12.5, and it is the counterpart of §17.13.

24. Indemnification

In short: If what you post or do gets us sued, you cover it — including our legal fees — and that obligation outlives your account.

24.1 Your indemnity. You will defend, indemnify, and hold harmless ACT 3 AI, Inc. and its officers, directors, employees, contractors, and agents.

24.2 What it covers. Every claim, demand, action, loss, damage, judgment, fine, penalty, settlement, and cost, including reasonable attorneys' fees and costs of investigation. This obligation is not limited by §22.3 (§22.14).

24.3 What triggers it. Claims arising out of or relating to: (a) your User Content; (b) your use of the Service; (c) your breach of these Terms or of any representation or warranty in them; (d) your violation of any law or of any right of any third party; or (e) your breach of an AI Provider's usage policy through your use of an AI feature (§18.14).

24.4 Specifically included. For the avoidance of doubt, §24.3 includes claims for defamation and related torts (§11.5), intellectual property infringement (§12), right of publicity, and privacy claims arising from your User Content — including where the content in question is AI Output you prompted, edited, adopted, or published (§18.11, §18.16).

24.5 We control our own defense. We may assume exclusive control of the defense and settlement of any indemnified claim at your expense. You may not settle any claim in a way that imposes an obligation on us, or admits fault on our behalf, without our prior written consent.

24.6 Notice and cooperation. We will give you reasonable notice of a claim for which we seek indemnity, and you will cooperate with the defense.

24.7 Survival. This section survives termination of your Account and of these Terms — the claims it covers, especially defamation and infringement claims, routinely arrive long after an Account is closed.

25. Dispute Resolution

In short: Complain to us, then send a Notice, then mediate in King County — all with real deadlines that run against us too, all deemed satisfied after 120 days, and none of it applying to small claims, to EU complaint routes, or to going to a regulator.

⚠️ SECTION 25 AFFECTS HOW DISPUTES BETWEEN US ARE RESOLVED, AND IT REQUIRES YOU TO TAKE SPECIFIC STEPS BEFORE STARTING ANY PROCEEDING. READ IT CAREFULLY — AND READ §25.13 FIRST, BECAUSE SEVERAL THINGS ARE CARVED OUT OF IT ENTIRELY.

25.1 Step two — the Notice of Dispute

Before starting any proceeding, and after completing §25.11, you must send us a written Notice of Dispute at the address in §30, containing the items listed at §25.15. We must do the same before proceeding against you. The parties will then attempt in good faith to resolve the dispute for 30 days from the date the Notice is sent, and each of us will participate personally as §25.16 requires. This is a condition precedent, subject to §25.12, §25.13 and §25.14.

25.2 Courts, not arbitration

A dispute that survives §25.12's ladder is heard in the courts identified in §26.2.

There is no arbitration agreement in these Terms, and no class-action waiver. We are saying that plainly rather than leaving you to infer it from silence, because most services you use do require arbitration and do waive class actions. You keep your right to a court, your right to a jury where one is available, and your right to participate in a class or representative proceeding. We do not ask you to give any of those up as a condition of using the Service.

If we ever decide otherwise, it takes a published amendment under §28.2 with its notice period — never a silent change — and §25.3, §25.8, §25.9 and §25.19 are reserved against that possibility and are not in force today.

25.3 Class-action waiver

(Reserved — not in force.) These Terms contain no arbitration agreement (§25.2). This number is held so that section numbering continues to match important_terms.md, which is how coverage is verified.

25.4 Small claims

Either party may bring an individual claim in small claims court, if it qualifies, without completing any step in this section (§25.13).

25.5 Injunctive relief

We may seek injunctive or other equitable relief in any court of competent jurisdiction to protect our intellectual property, the security of the Service, or confidential information, without completing any step in this section (§25.13). Public injunctive relief is different and is governed by §25.18.

25.6 Fee shifting

Each party bears its own attorneys' fees except where a statute or a court provides otherwise. We do not seek to shift our fees onto consumers, and any fee provision elsewhere in these Terms is subject to this sentence and to §25.20.

25.7 Contractual limitations period

ANY CLAIM OR CAUSE OF ACTION ARISING OUT OF OR RELATING TO THESE TERMS OR THE SERVICE MUST BE FILED WITHIN ONE (1) YEAR AFTER IT ARISES, OR IT IS PERMANENTLY BARRED — subject to the tolling in §25.14, and except where the applicable law of your jurisdiction does not permit shortening the limitation period, in which case the shortest permitted period applies.

25.8 Opt-out, and it is mutual

25.9 Mass-filing protocol

25.10 Step three — mandatory mediation in King County, Washington

(a) The obligation. After completing §25.11 and §25.1, and as a condition precedent to commencing any lawsuit, arbitration, or other formal proceeding against us, you and we agree to submit the dispute to non-binding out-of-court mediation before a neutral mediator. The parties may proceed to a formal proceeding only if the mediation does not resolve the dispute, or if §25.12's ceiling or failure provisions apply.

(b) Where. The mediation is seated in King County, Washington, United States of America. It may be conducted entirely by videoconference or telephone at your election, and you are never required to travel to Washington to satisfy this section. If you are a consumer resident in the European Union, the EEA, the United Kingdom, Switzerland, Norway, or Iceland, this section does not require you to mediate outside your own country and does not displace the courts of your country of residence (§25.13, §26.11).

(c) When it starts and when it ends. Either party may initiate by written notice to the other. It must be initiated within 30 days after the §25.1 period ends, and it is deemed concluded 45 days after initiation, whether or not a session has taken place. If neither party initiates within that window, this condition is deemed satisfied and both parties are free to proceed.

(d) Who pays. Where you are an individual consumer, we pay the mediator's fee in full. Each party otherwise bears its own costs, including its own lawyer if it brings one. You are not required to bring a lawyer, and §25.16 requires a decision-maker from our side to attend in person whether or not either of us brings one.

(e) Non-binding, and confidential. Mediation does not decide anything. It does not waive a jury trial, does not waive a class action, and is not arbitration (§25.2). What is said in mediation is confidential and inadmissible in any later proceeding to the extent the law allows, except for facts otherwise discoverable.

(f) This is our dispute with you, not your dispute with someone else. §25.10 applies to a dispute between you and us. It does not create any duty on our part to mediate a dispute between you and another Citizen or any third party — §23.3 governs that, and the two provisions are consistent.

25.11 Step one — you use our internal complaint process first

(a) The obligation. If you hold or have held an Account, you agree to file a complaint under §13.13 and to let that process run to its conclusion before taking any other step, including sending a Notice of Dispute, requesting mediation, filing suit, or filing a chargeback (§15.13). This is a condition precedent.

(b) It has a hard end. The process is over no later than 30 days after you file it (§13.13(c)), and it costs you nothing (§13.13(d)). It cannot be extended by us.

(c) It cannot be used to run out your clock. All limitation periods are tolled while it runs (§25.14), and it is subject to the ceiling and the failure provisions in §25.12.

(d) It does not replace anything. See §25.17. This process supplements your other routes; it does not substitute for them.

25.12 THE LADDER — THREE STEPS, WITH A HARD 120-DAY CEILING

⚠️ BEFORE YOU START ANY PROCEEDING AGAINST US, YOU AGREE TO COMPLETE THE FOLLOWING STEPS. EACH ONE IS A CONDITION PRECEDENT. READ §25.13 FIRST — SEVERAL THINGS ARE CARVED OUT OF IT ENTIRELY.

Step What Deadline
1 Internal complaint — §25.11, run under §13.13 We answer within 30 days; the step ends at 30 days no matter what
2 Notice of Dispute and good-faith discussion — §25.1 30 days from the Notice
3 Mediation in King County, Washington — §25.10 Must be initiated within 30 days of step 2 ending, and is over 45 days after it is initiated

THE CEILING. Every condition precedent in this section is deemed fully satisfied 120 days after you first file a complaint under §13.13 or send a Notice of Dispute under §25.1, whichever is earlier — whatever stage the steps have reached and whatever the reason. After that date you are free to proceed, and we will not argue otherwise.

IF WE DROP THE BALL, THE CONDITION IS SATISFIED IMMEDIATELY. Every condition precedent in this section is deemed satisfied at once if we: miss any deadline in §13.13 or in this section; fail to attend as §25.16 requires; decline or fail to participate in mediation; fail to pay our share of the mediator's fee under §25.10(d); or tell you in writing that we consider the process concluded.

THESE STEPS COST YOU NOTHING BEYOND YOUR OWN TIME. We do not charge for the complaint process (§13.13(d)) and we pay the mediator's fee for a dispute with a consumer (§25.10(d)).

THE REMEDY FOR NON-COMPLIANCE IS A STAY, NOT A FORFEITURE. If either party starts a proceeding before completing these steps, the other party's remedy is to ask the court or tribunal to stay that proceeding until the steps are complete or the 120 days have run. No claim is barred, waived, or lost by non-compliance with this section, and nothing here shortens any limitation period (§25.14).

Step	What	Deadline
1	Internal complaint — §25.11, run under §13.13	We answer within 30 days; the step ends at 30 days no matter what
2	Notice of Dispute and good-faith discussion — §25.1	30 days from the Notice
3	Mediation in King County, Washington — §25.10	Must be initiated within 30 days of step 2 ending, and is over 45 days after it is initiated

25.13 WHAT THE STEPS IN §25.1, §25.10, §25.11 AND §25.12 DO NOT APPLY TO

None of §25.1, §25.10, §25.11 or §25.12 applies to any of the following. These are carved out expressly and by name, because a condition precedent that swallowed them would not survive review — and would take the rest of the section down with it.

Small-claims court (§25.4). You may go straight there.

Injunctive or emergency relief sought by either party for intellectual property, security, safety, or confidentiality (§25.5), and public injunctive relief (§25.18).

EU and EEA recipients contesting a moderation decision. The Digital Services Act gives you an internal complaint-handling route (Article 20) and the right to take the decision to a certified out-of-court dispute settlement body (Article 21) — and Article 21 is expressly without prejudice to your right to initiate proceedings before a court at any stage. Nothing in §25 conditions, delays, or removes either of those rights (§26.4).

Consumers resident in the EU, the EEA, the United Kingdom, Switzerland, Norway or Iceland, to the extent the mandatory law of your country gives you a forum in your own country (§26.11).

The United Kingdom and any other jurisdiction whose law gives a comparable non-excludable route to a court, an ombudsman, or a statutory complaints body.

Any right that a mandatory consumer-protection law does not permit us to condition (§26.6).

Complaining to a regulator, a supervisory authority, a data protection authority, an attorney general, or law enforcement. You never have to talk to us first, and we will not treat it as a breach.

Reporting a crime, or a mandatory report we are required to make (§14.5).

25.14 Tolling

Every limitation period, contractual or statutory, is tolled from the moment you file a complaint under §13.13 or send a Notice of Dispute under §25.1, until the earlier of the completion of §25.12's steps or the expiry of its 120-day ceiling. Neither party may argue that time spent complying with this section has run against the other. Any filing-fee deadline is tolled on the same basis.

25.15 What a complaint or Notice of Dispute has to say

A complaint under §13.13 or a Notice of Dispute under §25.1 must include:

your name and an email address we can reach you at;
an express statement that you are invoking this process — the internal complaint process, or the Notice of Dispute, whichever you are sending;
the jurisdiction where you live (city and state, province, or country);
your Account handle, if you have one;
what happened, and when;
the specific relief you want — including the specific amount, if you are asking for money;
the supporting evidence you have, and a description of, or link to, any evidence you know of but do not hold.

That list is a floor, not a hurdle. If you send us those things, the requirement is satisfied — we will not refuse a complaint for form. We ask for them because a process that cannot identify what you want or what happened cannot resolve anything, and because a condition precedent with no defined contents cannot fairly be enforced against you either. The same requirements bind us when we send you a Notice.

25.16 Both of us turn up, in person

You and we each agree to participate personally in the §25.1 discussion and in the §25.10 mediation, by telephone or videoconference.

If you are represented by a lawyer, your lawyer may take part — but you agree to participate personally as well. Likewise, if we are represented by a lawyer, our lawyer may take part, but we will have a representative of ACT 3 AI, Inc. with authority to resolve the dispute participate personally.

This is symmetrical on purpose. A pre-suit process that requires a consumer to show up while the company sends only outside counsel is not a good-faith resolution step, and we do not want one. If we fail to attend as this section requires, every condition precedent in §25.12 is deemed satisfied immediately and you may proceed.

25.17 These steps supplement your rights; they do not replace them

The procedures in this section supplement, but do not replace, your right to pursue court actions or other regulatory complaints.

We may decline to run the internal complaint process again where an identical dispute between us has already been resolved through it. That is the only limit on it, and it does not affect §25.12's ceiling or §25.13's carve-outs.

25.18 Public injunctive relief goes to a court, not to arbitration

Where either party seeks public injunctive relief — that is, injunctive relief whose primary purpose and effect is to prohibit unlawful acts that threaten future injury to the public generally — the entitlement to and extent of that relief must be litigated in a civil court of competent jurisdiction and not in arbitration. Litigation of any issue of public injunctive relief is stayed pending the outcome of the merits of any individual claim.

This section applies whether or not Alternative B is adopted, and it is not severable from it. An arbitration agreement that purports to waive public injunctive relief in all forums is unenforceable in at least one major jurisdiction, and a clause that takes the rest of §25 down with it is worth less than no clause at all.

25.19 If the class-action waiver fails, the arbitration agreement fails with it

25.20 We do not recover our fees from you unless the claim was frivolous

If we prevail in any proceeding covered by this section, we will pay our own attorneys' fees and costs and will not seek to recover them from you, unless the court or arbitrator determines that your claim was frivolous or brought for an improper purpose. If you prevail, you are entitled to an award of your attorneys' fees and expenses to the extent the applicable law provides for one. This section controls over any general fee provision elsewhere in these Terms (§25.6), and it does not affect our right to recover costs from you under §10.7, §15.13 or §15.16, which are not fee-shifting provisions.

26. Governing Law, Venue, and International Terms

In short: King County, Washington for venue. But if you are a consumer in the EU, EEA, UK, Switzerland, Norway or Iceland, your own country's law and courts apply instead — we do not try to drag you to Washington.

26.1 Governing law. These Terms are governed by the laws of the State of Washington, United States of America, and by applicable United States federal law, excluding conflict-of-laws rules and excluding the UN Convention on Contracts for the International Sale of Goods. §26.11 is a real exception to this for consumers in the EU, EEA, UK, Switzerland, Norway and Iceland, and §26.6 is a further one wherever mandatory consumer law says so.

A note on why this says Washington, since it used to say nothing. ACT 3 AI, Inc. is a Washington corporation (§1.1), and the venue for both mediation and court proceedings is King County, Washington (§25.10(b), §26.2). Electing Washington law is the only combination of those three that is internally coherent — a Washington company, sued in Washington, under some third state's law, is a document that makes work for lawyers and sense to nobody. Choosing the governing law is nonetheless a formally separate decision from where a company is incorporated, and we record it here as a decision that was made rather than one the drafting performed quietly.

26.2 Venue. Subject to §25 and to §26.11, the state and federal courts located in King County, Washington, United States of America have exclusive jurisdiction over any proceeding arising out of or relating to these Terms or the Service, and both parties consent to personal jurisdiction there — except where mandatory consumer law entitles you to sue, or requires you to be sued, in your own country or state of residence (§26.6, §26.11), and except for the small-claims carve-out in §25.4.

26.3 Regional variations. We do not pretend one document fits every country. Where the law of your country requires different or additional terms, those terms apply to you and prevail over anything inconsistent here. We publish regional variations alongside these Terms.

26.4 European Union and EEA. We comply with the GDPR (see the Privacy Policy) and, for the Hosted Service, with the Digital Services Act, including: notice-and-action mechanisms; a statement of reasons for each moderation decision affecting an EU recipient; an internal complaint-handling system under Article 20, free of charge, decided under the supervision of qualified staff and not solely by automated means; out-of-court dispute settlement under Article 21 before a certified body of your choice, whose decisions are non-binding and which you can find through your national Digital Services Coordinator or the European Commission's public list; and transparency reporting. §13.2's "no obligation to explain" does not apply where the DSA requires a statement of reasons, and §25's conditions precedent do not apply to an EU recipient contesting a moderation decision (§25.13). To be explicit: you may go to a court in your own country at any stage, and none of §25 delays that.

26.5 United Kingdom. We take account of duties under the Online Safety Act, including in how §4.10 age assurance operates. UK Citizens should also note §11.8: UK defamation law is materially less favorable to publishers than US law, and a claim brought there may succeed where the same claim would fail in the United States.

26.6 Mandatory consumer protections prevail. Nothing in these Terms removes a right you have under the mandatory consumer law of your country of residence, including rights of cancellation, refund, warranty, forum, and access to a regulator. Where such a law conflicts with these Terms, that law wins. This applies with particular force to §15.10, §15.24, §17.11, §22.3 and §25.

26.7 Export control and sanctions. You will comply with all applicable export control and sanctions laws (§4.7).

26.8 Language. Where these Terms are translated, the English version controls unless the mandatory law of your jurisdiction provides otherwise. If we ever publish a translation as equally authoritative, we will say so in that translation and name which version prevails on an inconsistency (§26.12).

26.9 Written for citizens of any country. This movement is built for a citizen of any country; the reference social contract is per-country and the machinery is universal. That a US state's law governs (§26.1) and that US courts hear disputes (§26.2) is a practical necessity of contracting, not a statement that the Service or the movement is only for Americans.

26.10 Venue, stated once more because it is easy to miss. Mediation is seated in King County, Washington (§25.10(b)) and court proceedings are heard in King County, Washington (§26.2). You may attend mediation remotely (§25.10(b)) and you may use your local small-claims court (§25.4). The governing law in §26.1 is Washington — a separate question from venue, now answered the same way; see §26.1's note. §26.11 is a real exception to all of this and you should read it if you are outside the United States.

26.11 If you are a consumer in the EU, EEA, UK, Switzerland, Norway, or Iceland — your own law and your own courts.

If you are a consumer and you are a resident or citizen of any European Union country, or of the United Kingdom, Switzerland, Norway, or Iceland, then the governing law and the forum are the laws and the courts of your usual place of residence, and not §26.1 or §26.2. We may bring proceedings against you only in the courts of your country of habitual residence.

Nothing in §25 requires you to mediate outside your own country, and §25.13 carves you out of the conditions precedent to the extent your own law does not permit us to impose them.

We name the countries rather than writing "where local law requires", because a clause that makes a consumer work out for themselves whether it applies to them is not much of a protection.

26.12 If these Terms are ever translated. These Terms are published in English. If we publish a translation, we will state in that translation whether it is a convenience translation or an equally authoritative version, and — if equally authoritative — which language prevails on any inconsistency. We will not publish a translation that is silent on the point, because a reader in another language is entitled to know which text they are actually bound by (§26.8).

27. Term, Suspension, Termination, and Survival

In short: Leave whenever you like, without telling us — but cancel your subscriptions first. If we shut the Service down, we refund the unused part of what you prepaid.

27.1 Term. These Terms begin when you first accept them (§1.2) and continue until terminated.

27.2 You may leave at any time. You may stop using the Service and delete your Account at any time, from inside the Service. Cancel your Subscriptions before deleting your Account (§15.20, §27.9).

27.3 We may suspend or terminate. We may suspend or terminate your Account or access, with or without notice, for: breach of these Terms; conduct under §9; legal risk to us or to another person; or at our discretion under §8.16. Where we reasonably can, we will tell you why (§13.9).

27.4 Effects of termination. On termination: your licence under §6.1 ends; access to paid features ends and no refund is due if termination was for breach (§13.7); One-Time Purchases and Rentals on WeTheCitizens.tv end (§16.10); your local files are unaffected — they are on your machine and always were (§5.1); and published content is handled under §20.5 and §7.10, subject to the federation limits in §5.7 and §19.3.

27.5 Export window. Except where terminating for serious breach, unlawful conduct, or a legal requirement, we will give you at least 30 days to export your data before deletion (§20.4).

27.6 Discontinuing the Service. We may discontinue the Service or any feature. If we discontinue the Hosted Service entirely, we will give reasonable advance notice, provide an export window, and refund the unused portion of any prepaid Subscription (§27.8). This is a refund we commit to, and §15.24 does not switch it off. It does not extend to Donations (§17.11).

27.7 Survival. The following survive termination: §5 (local-first split), §7 (content licence, as limited by §7.4 and §7.10), §10 (accuracy, including §10.6), §11 (defamation), §12 (intellectual property), §14 (legal role), §15 (amounts already owed), §20 (privacy cross-reference), §21 (disclaimers), §22 (liability), §23 (release and covenant, including §23.6 and §23.7), §24 (indemnity), §25 (disputes, including the conditions precedent), §26 (governing law and venue), §29 (general), and this §27.7.

27.8 The refund we do commit to, stated as its own term. If we discontinue the Hosted Service entirely (§27.6), or if we make a material change to these Terms that you do not accept (§28.2), we will refund the unused prepaid portion of any Subscription on a pro-rata basis. This is a commitment we make freely; it is not something the law compels. It is expressly not switched off by §15.10, §15.24 or §17.11 — those are about refunds you request, and this is a refund we volunteer when we are the party changing the deal. It does not extend to Donations (§17.11) or to One-Time Purchases you have already substantially used (§16.8).

27.9 Leaving is meant to be easy. You can end this agreement at any time by deleting your Account. You do not need to notify us, ask our permission, give a reason, or speak to anyone.

The one thing to do first: cancel each Subscription (§15.2, §15.4). Deleting your Account does not cancel your Subscriptions, and billing can continue after you are gone (§15.20). We put those two sentences next to each other on purpose — a user who assumed deletion ended the billing is exactly how a chargeback happens, and neither of us wants that.

Some things survive your departure: §27.7 lists them, your published content on the Federated Network is beyond recall (§5.7), and content in our backups is governed by §7.10.

28. Changes to These Terms

In short: We give 30 days' notice of material changes, we explain why, and if you are a paid subscriber and you do not accept, you get the unused part of your money back.

28.1 We may change these Terms — within limits we accept now. We will post the updated version with a new effective date and version number, and notify you in the Service or by email. We will make changes only where these Terms have become inappropriate or incomplete for the Service as it actually is, or where a change is required for security or to comply with law, and only where the change is reasonable for you, taking your interests into account. We will tell you what changed and why (§28.6).

28.2 Material changes get advance notice. For material changes we will give at least 30 days' notice before they take effect, together with an explanation of the change and the reason for it. If you hold a paid Subscription, you may cancel before the change takes effect and receive a refund of the unused prepaid portion if you do not accept it (§27.8). This is a refund we commit to, and §15.24 does not switch it off.

28.3 Continued use is acceptance. Continuing to use the Service after the effective date of a change is acceptance of the changed Terms. If you do not accept, stop using the Service and delete your Account (§27.2, §27.9).

28.4 Version history. We keep a public change log showing what changed, when, and why, and we keep prior versions available.

28.5 No retroactive effect. Changes do not apply retroactively to a dispute that arose before the change took effect.

28.6 A limit on the amendment right itself, not only on the notice. Most contracts let one party rewrite them at will and offer only a notice period in return. We are also limiting what kind of change we may make at all, and §28.1 states that limit: a change must be needed because these Terms have stopped fitting the Service, or because security or law requires it, and it must be reasonable for you with your interests taken into account. We will explain the change and the reason for it, not merely announce that something changed. We accept that limit because a unilateral amendment right with no substantive constraint is one of the least defensible things in any consumer contract, and because we would rather not have one.

29. General Provisions

In short: The usual machinery — severability, assignment, notices — plus two unusual ones: our "board of directors" binds nobody (§29.12), and government users get their own rules (§29.14).

29.1 Entire agreement. These Terms, together with the documents in §1.5, are the entire agreement between us about the Service and supersede any prior understanding. Nothing in this section limits liability for fraudulent misrepresentation (§22.6, §22.12, §29.13).

29.2 Severability, with reduction preferred. If any provision is held unenforceable, the parties intend that it be reduced or reformed to the minimum extent necessary to make it enforceable, rather than struck out, and the remaining provisions continue in full force. This applies with particular force to §10.6 (see §10.6(d)), §22.3 (see §22.10), §22.13, §23.2, §23.6 and §25. The one deliberate exception is §25.19, which voids rather than reduces, for the reason stated there.

29.3 No waiver. Our failure or delay in enforcing any provision is not a waiver of it, and a waiver on one occasion is not a waiver on another.

29.4 Assignment. We may assign these Terms, in whole or part, to an affiliate or in connection with a merger, acquisition, reorganization, or sale of assets. You may not assign them.

29.5 No third-party beneficiaries. There are no third-party beneficiaries of these Terms, except that the persons indemnified under §24.1 may enforce §24. In particular, a person who is not a registered User acquires no right under these Terms (§3.6) — which does not affect their ability to use the reporting route at §11.6, a right we give them outside this contract.

29.6 Force majeure. Neither party is liable for a failure to perform caused by an event beyond its reasonable control. This does not excuse an obligation to pay.

29.7 Relationship. We are independent parties. Nothing here creates an agency, partnership, joint venture, employment, or fiduciary relationship.

29.8 Headings. Headings, the Table of Contents, and the "In short" lines are for convenience only and do not affect interpretation (§1.8).

29.9 Notices. We may notify you in the Service, or at the email address on your Account. You notify us at the address in §30. Notice is effective when sent, unless a law requires otherwise. Keep your contact details current (§4.2).

29.10 Electronic contracting. You consent to contract electronically and to receive records, notices, and disclosures electronically. You may withdraw that consent, but doing so may mean you can no longer use the Service.

29.11 US Government end users. The software is "commercial computer software" under FAR 12.212 and DFARS 227.7202, licensed to US Government end users only with the rights granted to all other users. See §29.14, which is broader and governs official government use of the Service.

29.12 The advisory group binds nobody. Members of the advisory group defined in §2.8 — however the Service labels them, including as a "board of directors" — hold no corporate office at ACT 3 AI, Inc., owe you no fiduciary duty or duty of care, and have no actual, apparent, or ostensible authority to bind, represent, contract for, or speak on behalf of ACT 3 AI, Inc. Their views are their own. Nothing an advisory group member says or writes is a representation, warranty, or commitment by us, and nothing in §2.8 or this §29.12 creates a partnership, joint venture, agency, or employment relationship (§29.7).

29.13 Fraud is carved out of the entire-agreement clause. Nothing in §29.1, and nothing elsewhere in these Terms, excludes or limits any liability or remedy either party has for fraudulent misrepresentation or fraudulent concealment. An entire-agreement clause that purported to do that would be unenforceable in most jurisdictions and would put §29.1 itself at risk, so we state the carve-out rather than rely on it being implied (§22.6, §22.12).

29.14 Government entities and officials acting officially.

This section applies where a federal, state, provincial, tribal, or local government entity, or an official acting in an official capacity, uses the Service for official purposes. It controls over anything inconsistent elsewhere in these Terms.

Indemnification. §24 does not apply to your official governmental use, except to the extent your applicable law authorizes it.

Governing law, venue, and dispute resolution. §25 and §26 do not apply to the extent you are legally prohibited from agreeing to them. For a United States federal entity, liability is determined under the Federal Tort Claims Act or other governing authority, and federal limitation periods apply. For other government entities, the law and forum your own authority requires apply instead.

The covenant not to sue. §23.6 binds an individual personally and does not bind an office (§23.7).

Litigation conduct. Any provision of these Terms regulating the conduct of litigation is subject to the legislative requirements that apply to your official use.

No implied endorsement, either way. We will not use your seal, insignia, or name to imply that you endorse us (§23.8), and your use of the Service does not imply that we endorse you (§17.4, §17.13).

Everything else in these Terms continues to apply, including §8 (acceptable use), §10 (accuracy), §11 (defamation), and §12 (intellectual property). If any provision, whether or not modified by this section, is inconsistent with the law that governs your use, tell us at the address in §30 and we will discuss an alternative rather than leave you bound by something you cannot lawfully accept.

30. Contact and Legal Notices

In short: Where to send everything. Two of these addresses — complaints and corrections — must be live before this document is published, because other sections point at them.

Purpose	Contact
General contact — the address to use if none of the rows below fits	ContactUs@ACT3ai.com
Legal entity and notice address	ACT 3 AI, Inc., a Washington corporation, doing business as We the Citizens — 16430 NE 50th Street, Redmond, WA 98052, United States of America
Legal notices, Notice of Dispute (§25.1), mediation notice (§25.10), arbitration opt-out (§25.8)	ContactUs@ACT3ai.com (subject line: "Legal notice")
Complaints about us — the internal complaint process (§13.13, §25.11)	ContactUs@ACT3ai.com (subject line: "Complaint")
Report a mistake — corrections in content or AI Output (§13.12, §18.12, §21.9)	ContactUs@ACT3ai.com (subject line: "Correction")
Abuse, harassment, and safety reports (§13.3)	ContactUs@ACT3ai.com (subject line: "Abuse")
Reports of false statements about you (§11.6) — no Account required	ContactUs@ACT3ai.com (subject line: "Abuse")
DMCA designated agent (§12.7–§12.9)	ContactUs@ACT3ai.com (subject line: "DMCA") — or by post to ACT 3 AI, Inc., Attn: Copyright Agent, 16430 NE 50th Street, Redmond, WA 98052, United States of America. Not yet registered with the U.S. Copyright Office (§3.7, §14.4).
Trademark complaints (§12.12) — NOT the DMCA agent; the DMCA does not cover trademark	ContactUs@ACT3ai.com (subject line: "Legal notice")
Law-enforcement requests (§13.6)	ContactUs@ACT3ai.com (subject line: "Law enforcement")
Government entities and officials (§29.14)	ContactUs@ACT3ai.com (subject line: "Legal notice")
Privacy and data-subject requests	See the Privacy Policy
EU/UK representative, where required (§26.4–§26.5)	Not applicable - we do not offer the Service to people in the European Economic Area or the United Kingdom, and we do not monitor their behaviour, so Article 27 does not require us to appoint a representative

Last modified: 2026-08-19 · Version: 0.6 · Change log: see §28.4 · Previous versions: 0.5 · 0.4 · 0.3 · 0.2 (all 2026-08-19) · 0.1 (2026-08-18)

The Short Version​

Table of Contents​

1. The Agreement and Acceptance​

2. Definitions​

3. Scope — One Product, Many Front Doors​

4. Eligibility, Accounts, and Identity​

5. The Local-First Split — What Lives Where​

6. License to Use the Service​

7. User Content and the License You Grant Us​

8. Acceptable Use​

9. Inauthentic Accounts and Coordinated Narrative Control​

10. Data Accuracy and Integrity​

11. Defamation and Statements About Real People​

12. Intellectual Property, Infringement, and DMCA​

13. Moderation, Enforcement, and Appeals​

14. Our Legal Role — Safe Harbors​

15. Payment, Credit Cards, and Subscriptions​

16. Streaming, Pay-Per-View, and Rentals (WeTheCitizens.tv)​

17. Donations, Fundraising, and Campaign Finance​

18. AI Features and AI Output​

19. Third Parties, Federation, and Integrations​

20. Privacy and Data Handling​

21. Disclaimers of Warranty​

22. Limitation of Liability​

23. Release and Covenant Not to Sue​

24. Indemnification​

25. Dispute Resolution​

25.1 Step two — the Notice of Dispute​

25.2 Courts, not arbitration​

25.3 Class-action waiver​

25.4 Small claims​

25.5 Injunctive relief​

25.6 Fee shifting​

25.7 Contractual limitations period​

25.8 Opt-out, and it is mutual​

25.9 Mass-filing protocol​

25.10 Step three — mandatory mediation in King County, Washington​

25.11 Step one — you use our internal complaint process first​

25.12 THE LADDER — THREE STEPS, WITH A HARD 120-DAY CEILING​

25.13 WHAT THE STEPS IN §25.1, §25.10, §25.11 AND §25.12 DO NOT APPLY TO​

25.14 Tolling​

25.15 What a complaint or Notice of Dispute has to say​

25.16 Both of us turn up, in person​

25.17 These steps supplement your rights; they do not replace them​

25.18 Public injunctive relief goes to a court, not to arbitration​

25.19 If the class-action waiver fails, the arbitration agreement fails with it​

25.20 We do not recover our fees from you unless the claim was frivolous​

26. Governing Law, Venue, and International Terms​

27. Term, Suspension, Termination, and Survival​

28. Changes to These Terms​

29. General Provisions​

30. Contact and Legal Notices​