ComplianceEffective date1 Jan 2026US
California requires companion chatbots to disclose they are AI and run self-harm protocols
260days
Until July 1, 2027
First annual report to the Office of Suicide Prevention due
Since 1 Jan 2026, a companion chatbot offered to California users has to tell people it is AI when they could think it is human, and remind known minors at least every three hours. Operators keep and publish a protocol for suicide and self-harm that refers users to crisis services, and block sexual content aimed at minors. Injured users can sue for at least $1,000 per violation, and annual reports to the Office of Suicide Prevention start on 1 Jul 2027.
Build: Publish the self-harm protocol, add AI disclosures and three-hour reminders for minors, and start logging crisis referrals for the 2027 report.
Who it reaches
It reaches generative AI services with users in the US, at any size. Operators offering a companion chatbot to California users. Customer service, business and game-only bots are excluded.
Lands on Policy & Standards, Product, Engineering, Legal & Regulatory and Data & Analytics.
Learn more
Read the sourceSkadden
Not legal advice. Read the source.