SB 243
In plain wordsA 2025 California law requiring AI companion chatbots to tell users they are not human, respond safely when people talk about suicide, and protect minors, including with regular reminders to take a break.
California's law, in force since January 1, 2026, on companion chatbots: conversational AI that adapts to its user and behaves enough like a person to keep up an ongoing relationship and meet social needs. Operators must clearly disclose that the chatbot is artificial where a reasonable person could be misled into thinking it human; maintain and publish protocols to prevent content promoting suicidal ideation or self-harm, including referring users who express such thoughts to crisis services; and, for users known to be minors, disclose that they are talking to AI, remind them at least every three hours to take a break and take reasonable measures to block sexually explicit content. From July 2027 operators must report annually to the state Office of Suicide Prevention. People injured by violations may sue for damages of at least $1,000 per violation. Customer-service bots, limited video-game characters and voice assistants are excluded.
To comply with SB 243, the companion app now shows minors a break reminder every three hours.
Origin
Introduced by State Senator Steve Padilla after reports of teenagers harmed in relationships with companion bots, including the 2024 suicide at issue in Garcia v. Character Technologies; signed by Governor Gavin Newsom on October 13, 2025, the same day he vetoed a broader bill restricting minors' access to companion chatbots.
Usage note
Often described as the first US state law aimed specifically at the safety of AI companionsAI companion, n.: A chatbot designed to be a friend, partner or confidant rather than a tool: it remembers you, has a personality and keeps up an ongoing relationship.; China's 2026 measures on human-like interactive AI services impose comparable duties, including special protections for minors.
Related
- AI companionAI companion, n.: A chatbot designed to be a friend, partner or confidant rather than a tool: it remembers you, has a personality and keeps up an ongoing relationship.
- AI liabilityAI liability, n.: Who has to pay or answer in court when an AI system causes harm — the company that built it, the business that used it or the person who relied on it — and under which laws.
- chatbotchatbot, n.: Software you chat with in ordinary language, by typing or talking, as you would with a person. The newest ones, such as ChatGPT, can answer questions, draft writing and keep long conversations going.
- SB 53SB 53, prop. n.: A 2025 California law requiring developers of the most powerful AI models to publish their safety plans, report serious safety incidents to the state and protect employees who raise alarms.
- AI preemptionAI preemption, n.: The push in the United States for a single national AI law that would override the states' own AI laws, and the fight over whether that is wise or even lawful.