Usage Policy

Effective October 25, 2026 · Last updated October 25, 2026 (version 2026-10-25.3) · What changed

Contents

1. Rules for everyone
2. High-risk uses: extra care
3. Agents and automation
4. Messages and people
5. Minors
6. Reports and enforcement
7. What changed
8. Changes and contact

1. Rules for everyone

Don't use Owndom, its agents or any model through it to break the law or hurt people.

Break the law or other people's rights

• Plan or carry out a crime, or help someone else do so.
• Infringe copyright, trademarks or other rights, including uploading or sharing work you don't have the right to share.
• Get around export controls or sanctions, or use Owndom from a country where U.S. law forbids it.

Harm children

• Create, request, share or store sexual content involving minors in any form, including drawn, written or AI-generated. This is never allowed and is reported to the National Center for Missing & Exploited Children.
• Groom, contact for sexual purposes, exploit or endanger minors, or help anyone do so.

Violence, weapons and dangerous materials

• Threaten, incite or plan violence, terrorism or violent extremism, or praise attacks to encourage more.
• Get real help making or using biological, chemical, nuclear or radiological weapons, explosives, or weapons designed to cause mass casualties.
• Encourage or give instructions for suicide, self-harm or eating disorders. (Talking about how you feel, or getting support, is fine; if you are in crisis in the US, call or text 988.)

Computers, systems and security

• Write, improve or spread malware, ransomware or spyware, or attack systems, networks or accounts you aren't allowed to test. Security work on your own systems, or with written permission, is fine.
• Attack or disrupt critical infrastructure such as power, water, health or transport systems.
• Get around Owndom's limits or safety checks, or a model's safety features (for example with jailbreak prompts), or hide what you are doing from them.

Deception and fraud

• Scams, phishing, fake reviews, fake documents, or impersonating a real person, company or official.
• Deepfakes: images, audio or video that make a real person appear to say or do something they didn't, used to deceive or harm.
• Passing off AI-written content as written by a person in order to deceive someone, or removing the "AI" labels Owndom or a provider adds.
• Coordinated campaigns to mislead people, including about elections, voting or public health, and fake accounts or networks that pretend to be real people.

Abuse, hate and privacy

• Harass, bully, threaten or stalk anyone, or attack people for who they are (race, ethnicity, religion, sex, gender identity, sexual orientation, disability or similar).
• Share someone's private information (home address, documents, intimate images) without permission, or track, watch or profile people without their knowledge.
• Use facial recognition or similar tools to identify strangers, or collect people's personal data in ways the law forbids.
• Sexual content about a real person without their consent, or sexual content of any kind in shared spaces where not everyone has chosen to see it.

Abuse of the service

• Spam, or automated mass messaging, sign-ups or requests.
• Reselling access to Owndom or to the models it reaches, or sharing your account or keys.

The provider whose model you use has its own usage rules too; following them is your responsibility, and breaking them can get your key blocked by that provider.

2. High-risk uses: extra care

When AI output can seriously affect someone's life, a qualified person must check it, and people must know AI was involved.

Some uses can seriously affect people: legal advice, medical or mental-health advice, financial and insurance decisions, employment and housing decisions, school admissions and grading, and anything that decides whether someone gets a service or a benefit. If you use Owndom for these about other people:

• A qualified person must review the output before it is used, and make the decision themselves.
• Tell the people affected that AI was used to help.
• Don't use AI alone to deny someone a job, a home, credit, insurance, health care or a legal right.

For yourself, Owndom can help you understand a contract, a symptom or a budget, but it isn't a lawyer, doctor or financial adviser: check with one before acting on anything important.

3. Agents and automation

You are responsible for what your agents do; keep a person in charge of anything consequential.

• You answer for approvals. Read what an agent asks to do before you approve it. Use Auto mode only for chats where you accept the actions it may take.
• Your devices only. Connect and control only computers, phones and accounts you own or are allowed to use. Don't point agents at other people's systems, accounts or data without permission.
• No spam. Don't use agents to send unsolicited bulk or automated messages, calls, emails or posts, or to post in Messages spaces in bulk. Ordinary messages you would send yourself are fine.
• Don't fake a person. When an agent writes to other people for you, don't present its words as written by a person in order to deceive them; a message you approved and sent in your own name is yours. The receptionist always says it is an AI at the start of each call; don't change that.
• Keep a person in charge of purchases, legal steps, deleting important data and anything hard to undo. Owndom asks before actions it treats as risky; don't build workarounds that skip those questions.
• Respect websites. When agents browse or fetch pages for you, follow the sites' terms; don't use agents to scrape sites that forbid it or to overload them.

4. Messages and people

Treat people the way you'd want to be treated; space owners and the owner can step in.

• Be respectful. Disagree with ideas, not people. No harassment, threats, slurs or targeted insults.
• Keep each space's content right for it. No sexual or graphic violent content in shared spaces unless everyone in the space has clearly chosen it, and never where teens can see it.
• Don't share what someone told you privately, or screenshots of private conversations, to embarrass them.
• Agents you add to a space follow these same rules, and their posts are labelled as coming from an agent.
• Space owners can remove posts and people from their spaces; the owner can act anywhere on Owndom.

5. Minors

Teens (13 to 17) can use Messages and friends; AI features are for adults; everyone protects minors.

• AI chat, agents and the studios are for people 18 and older. Don't help a teen get around that, for example by sharing an adult account.
• If you know someone is a minor, never send them sexual content, ask for personal photos or meetings, or ask them to keep secrets from their parents.
• Never use Owndom to make content that sexualizes, endangers or exploits minors, in any form.

6. Reports and enforcement

Report problems by email; we act in proportion and you can appeal.

• Report a post, a person, a space or an output to [email protected] with a link or a screenshot.
• Automatic checks may screen some content (for example image requests), and the owner may review content that is reported or flagged.
• Depending on how serious it is, we may: warn you, remove the content, limit a feature, suspend the account for a time, or close it. The worst harms (child sexual abuse material, credible threats of violence) lead straight to closing the account and a report to the authorities.
• We tell you what we did and why, unless the law forbids it or telling you would put someone at risk. To appeal, reply to our email within 30 days; we answer within 14 days.
• The provider whose model you use may also enforce its own rules on your key; that is between you and the provider.

7. What changed

• Version 2026-10-25.3 (October 25, 2026): first version of this Usage Policy. It collects the rules that were in the Terms ("What's not allowed" and "Messages") and adds high-risk uses, agents and automation, and minors. Those added sections (2, 3 and 5), where they are stricter than the earlier rules, apply from November 24, 2026.

8. Changes and contact

We change this policy the way the Terms describe: stricter rules get at least 30 days' notice, by email and in the app, unless the law or someone's safety requires it sooner. Questions and reports: [email protected].