What are AI companion laws?
AI companion laws are rules written specifically for chatbots that act like a friend, partner or confidant and remember you across conversations. They are new: the first ones in the US took effect in late 2025, and more states have followed. Most of them ask for the same few things: tell people they are talking to AI, have a plan for when a user mentions suicide or self-harm, and add extra protections for minors.
If you are thinking about starting an AI girlfriend site, these laws matter because an AI girlfriend site is exactly the kind of product they describe. California's law, for example, defines a companion chatbot as an AI system with "adaptive, human-like responses" that can meet a user's social needs and "sustain a relationship across multiple interactions" (SB 243 bill text). That is a fair description of every companion on a typical AI girlfriend website.
This guide is general information for beginners, not legal advice. Laws change, and what applies to you depends on where you and your members are. Talk to a qualified lawyer before you launch.
Is an AI girlfriend business legal in the US?
Yes, running an AI girlfriend business is legal in the US, as long as you follow the rules that apply to it. No federal law bans AI companions for adults. What you do have is a growing set of state companion laws, age verification laws for sexual content, payment processor rules and general consumer protection law. Each companion must be an adult character, and the site must keep minors out.
Think of it like opening a bar. Selling drinks is legal, but you still need to check IDs, follow local rules and keep your payment provider happy. The legal risk in this business rarely comes from the idea itself. It comes from skipping the boring parts.
What does New York's AI companion law require?
New York was first. Its AI companion law, General Business Law Article 47, took effect on November 5, 2025, according to the Governor's office. Operators of AI companions must:
- Run a crisis protocol. Detect signs of suicidal ideation or self-harm and promptly point the user to crisis service providers.
- Remind users it is AI. Show a clear notice that the user is talking to AI, not a human, at the start of a session and every three hours of continued use.
The New York Attorney General enforces the law, and the Governor's office says fines collected go toward suicide prevention programs. Notice that the law is not limited to minors: the reminders apply to every user.
What does California's SB 243 require?
California's SB 243 was approved on October 13, 2025 and took effect on January 1, 2026, as TechCrunch reported. It applies to any "operator," which the bill text defines as a person who makes a companion chatbot platform available to a user in California. There is no size threshold in that definition, so a small site with California members is covered. The main duties:
- If a reasonable person could be misled into thinking they are talking to a human, show a clear notice that the companion is artificially generated.
- Keep a protocol that refers users who express suicidal thoughts or self-harm to a crisis line, and publish details of that protocol on your website.
- Tell users that companion chatbots may not be suitable for some minors.
- For a user you know is a minor, add break reminders at least every three hours and block sexually explicit material.
- Starting July 1, 2027, report crisis referral numbers and protocols each year to California's Office of Suicide Prevention.
SB 243 also lets a person injured by a violation sue for the greater of actual damages or $1,000 per violation, plus attorney's fees. That private right to sue is why lawyers pay close attention to it.
Which other states have AI companion laws?
Washington and Oregon both passed companion chatbot laws in March 2026, and both take effect on January 1, 2027, according to a Mayer Brown legal update. Washington's HB 2225 requires an AI disclosure for all users at the start of a chat and at least every three hours, every hour for known minors, and bars the companion from claiming to be human, even when asked. Oregon's SB 1546 requires a crisis protocol that refers users to the 988 Suicide and Crisis Lifeline, published on your website, plus stricter limits for users who may be minors.
Here is how the four laws compare at a glance:
| State | Law | In effect | Key duties | Who enforces |
|---|---|---|---|---|
| New York | General Business Law Article 47 | November 5, 2025 | Crisis protocol; "not a human" notice at start and every 3 hours | Attorney General |
| California | SB 243 | January 1, 2026 | AI notice; published crisis protocol; minor protections; annual report from 2027 | Private lawsuits ($1,000 per violation minimum) |
| Washington | HB 2225 | January 1, 2027 | AI notice every 3 hours (hourly for minors); no claiming to be human; crisis protocol | Consumer Protection Act, including private lawsuits |
| Oregon | SB 1546 | January 1, 2027 | AI notice; 988 referral protocol; minor protections | Private lawsuits ($1,000 per violation minimum) |
Other states are considering similar bills, so treat this table as a snapshot. A website can be reached from anywhere, which means the strictest law where your members live is the one to plan around.
Do you need age verification for an AI girlfriend site?
If your site offers adult content, very likely yes. In June 2025 the US Supreme Court upheld a Texas law that requires certain commercial websites publishing sexually explicit content to verify that visitors are 18 or older (Free Speech Coalition v. Paxton). Other states have passed age verification laws of their own, and the rules differ on which sites they cover and which verification methods count.
Payment processors add a second layer. When Mastercard tightened its adult content rules in 2021, it required the banks that connect adult merchants to certify that the seller has controls to monitor, block and take down illegal content, along with documented age and identity verification, content review before publication, and a complaint process that resolves reports within seven business days. In practice, the processor that approves your site will check all of this, and it is often the hardest step for first-time founders.
Do AI companion laws apply outside the US?
Yes. In the European Union, Article 50 of the AI Act applies from August 2, 2026. It requires providers to design AI systems that interact with people so that users are told they are dealing with AI, unless that is already obvious (AI Act, Article 50). If you plan to accept members from Europe, plan for that disclosure from day one. WhiteLabelFans covers the EU side in more depth for operators in its AI Act compliance checklist for adult AI operators.
Is the federal government regulating AI companions?
Not with a specific law yet, but it is watching. In September 2025 the Federal Trade Commission ordered seven companies, including Meta, OpenAI, Character Technologies and Snap, to explain how they test and monitor the effects of companion chatbots on children and teens, and how they monetize user engagement. The FTC describes these 6(b) orders as studies without a specific law enforcement purpose. Even so, it is a clear sign that regulators care most about minors, disclosure and how companion products make money. General consumer protection law, such as rules against deceptive advertising, already applies to every business.
What compliance checklist should a beginner follow?
Before you launch, make sure you can answer yes to each of these, whether you build the site yourself or use a provider:
- Every companion is an adult character, and none is based on a real person's likeness.
- The site clearly tells users they are chatting with AI, with repeat reminders during long sessions.
- There is a written protocol that refers users who mention suicide or self-harm to crisis resources, and it is published on the site.
- Visitors must verify they are adults before reaching adult content, using a method accepted where your members live.
- Your payment processor has approved the site and its content rules, including content review and a complaint process.
- Your terms of service, privacy policy and age notices are written for this category, ideally reviewed by a lawyer.
- Your promotion never suggests minors, never uses real people without permission and never promises results.
If you are still deciding how to get your site built, our pillar guide on how to start an AI companion business walks through each route, and the free startup cost calculator helps you budget for compliance work if you go it alone.
What does a done-for-you provider handle, and what stays with you?
On a done-for-you platform, the provider runs the technical side: the site, AI chat, billing, age checks and the content rules your processor requires. That removes most of the work in the checklist above. It does not remove all of your responsibility. You still choose your brand and niche, and you are responsible for how you promote it, including following each social platform's rules.
Whatever route you pick, ask the provider to show you how its platform handles AI disclosure, crisis referrals and age verification, and get that in writing. Our comparison of done-for-you and DIY AI businesses covers the other questions to ask, and the AI girlfriend business model guide explains how the money flows once you are live. For the questions beginners ask most, see our roundup of AI girlfriend business questions from Reddit.
Frequently asked questions
Is AI girlfriend legal?
Yes. Using or running an AI girlfriend site is legal in the US for adults. Operators must follow the rules that apply: state AI companion laws in places like New York and California, age verification laws for sexual content, payment processor rules and consumer protection law. Every companion must be an adult character. This is not legal advice.
What is the AI companion law in New York?
New York's General Business Law Article 47 took effect on November 5, 2025. AI companion operators must detect signs of suicidal ideation or self-harm and refer users to crisis services, and must tell users they are not talking to a human at the start of a session and every three hours. The Attorney General enforces it.
What is SB 243 in California?
SB 243 is California's companion chatbot law, in effect since January 1, 2026. It requires a clear AI notice when users could think they are talking to a human, a published crisis protocol, a notice that companions may not suit some minors, extra protections for known minors and, from July 1, 2027, annual reports to the state.
Do AI companion laws apply to small websites?
Often, yes. California's SB 243 covers anyone who makes a companion chatbot platform available to users in California, with no minimum size in that definition. Because a website can be reached from many states, plan around the strictest law where your members live, or use a provider that builds these features in.
Can you get sued under an AI companion law?
In some states, yes. California and Oregon let an injured user sue for the greater of actual damages or $1,000 per violation, and Washington's law is enforced through its Consumer Protection Act, which allows private lawsuits. New York's law is enforced by the Attorney General. Talk to a lawyer about your own situation.