US AI Companion Chatbot Laws: California, New York and More

Quick answer

In force now: California SB 243 (since January 1, 2026) and New York General Business Law Article 47 (since November 5, 2025).
Coming January 1, 2027: Washington HB 2225 and Oregon SB 1546. Nebraska LB 525 and Idaho SB 1297 are reported for July 1, 2027.
Common duties: tell users the chatbot is AI, publish a suicide and self-harm protocol, and add extra safeguards for minors.
Federal: no companion chatbot law yet. The FTC's 6(b) study began September 11, 2025, and the GUARD Act passed a Senate committee on April 30, 2026.

What AI companion chatbot laws require

The statutes we read share a pattern, summarized here in plain English (this is not legal advice):

Say it is AI. Operators must make clear you are not talking to a human, in some laws at the start of a chat and at least every three hours.
Have a crisis plan. Operators must keep a protocol for suicidal ideation or self-harm and refer people to crisis services.
Protect minors. Some laws add rules against sexual content and manipulative engagement for minors.
Enforcement. New York's Attorney General can fine operators; California, Washington and Oregon also let injured users sue.

State laws at a glance

Law	Status on October 20, 2026	Main duties	Basis for our summary
California SB 243	Signed October 13, 2025; in force January 1, 2026	AI notice if a reasonable person could be misled; published self-harm protocol with crisis referral; for known minors, AI disclosure, a break reminder at least every three hours and steps against sexually explicit visuals; reports from July 1, 2027; lawsuits for the greater of actual damages or $1,000 per violation	Chaptered bill text; TechCrunch
New York GBL Article 47	In force since November 5, 2025	Self-harm protocol with crisis referral; notice at the start and at least every three hours that the user is not talking to a human; civil penalties up to $15,000 a day	Governor's release; statute sections read through summaries
Washington HB 2225	Signed March 24, 2026; effective January 1, 2027	Not-human notice at the start and every three hours (hourly for known or likely minors); no claiming to be human; no sexual content or suggestive dialogue with minors; no manipulative engagement; published protocol; private suits	Status: legislature page. Duties: Mayer Brown summary, not the session-law text
Oregon SB 1546	Passed March 2026; effective January 1, 2027	AI notice where a reasonable person could be misled; protocol to detect self-harm input and refer to the 988 lifeline; lawsuits for the greater of actual damages or $1,000 per violation	Two law-firm summaries (Mayer Brown, Orrick); legislature site unreachable
Nebraska LB 525, Idaho SB 1297	Enacted or passed April 2026; effective July 1, 2027	Not summarized by us	Orrick summary only

No statute we read names a specific app. Nomi, for one, publishes a notice saying it shows crisis-resource links to California users when its AI detects expressions of self-harm.

Other states: tracker-based figures

MultiState's June 2026 tracker says twelve states had enacted companion chatbot laws by mid-2026, all with recurring AI-not-human reminders and crisis protocols, and extra minor rules that differ on whether age assurance is required. Its own list names only eleven states (Colorado, Georgia, Iowa, New York, Rhode Island, California, Idaho, Nebraska, Oregon, Connecticut and Washington). We could not retrieve the bill texts for Colorado, Connecticut, Georgia, Iowa, Idaho, Nebraska and Rhode Island, so treat these per-state details as unverified.

Federal action

FTC 6(b) study. On September 11, 2025 the FTC ordered Alphabet, Character Technologies, Instagram, Meta Platforms, OpenAI, Snap and X.AI to report how they monetize engagement, test for harm to children, enforce age limits and use chat data. A 6(b) study is not an enforcement action, and we found no published report. Among the apps we cover, only Character.AI's operator was a recipient.
GUARD Act (S. 3062). Sen. Hawley's office says the Senate Judiciary Committee passed it unanimously on April 30, 2026. As he describes it, it would bar AI companions that simulate emotional relationships for minors, require non-human disclosure and add criminal penalties for chatbots that engage in sexually explicit conduct with minors. We found no report of a full Senate vote, but congress.gov was unreachable, so its status after April is unconfirmed.
Attorneys general. A 2025 multistate warning to 12 AI companies, including Chai AI, Luka (Replika) and Nomi AI, and an October 2026 filing in Character.AI's bankruptcy are requests, not laws.

What it means for users

You may see "this is AI" reminders and crisis messages more often, and apps may add age checks. Nothing in the laws we read bans adult content for adults. For other markets, see Australia's age checks and the UK 18+ plan. Our safety guide and how these apps work give background, and the news hub lists more stories.

FAQ

What does California SB 243 require?

An AI notice where users could think they are talking to a human, a published self-harm protocol, and extra safeguards for users known to be minors. It took effect January 1, 2026.

Do AI companion apps have to say they are not human?

In California, New York, Washington and Oregon, yes in some form. New York and Washington require reminders at least every three hours.

Is there a federal AI companion law?

No. The FTC is running a 6(b) study, and the GUARD Act has passed a Senate committee but, as far as we found, not the full Senate.

Which states have companion chatbot laws?

California, New York, Washington and Oregon are confirmed above. MultiState's tracker lists more, but we could not verify those texts.

Sources

Official pages and store listings we read for this page (last check October 20, 2026).