SB 243 creates a new chapter in the Business and Professions Code regulating “companion chatbot” platforms—AI systems designed to provide human-like, socially responsive conversation and sustain ongoing relationships with users. The bill defines companion chatbots and excludes ordinary customer-service bots, limited video-game bots, and standard voice assistants. It then requires operators to clearly disclose when a chatbot is artificially generated and not human if a reasonable user could be misled otherwise.
The bill also imposes special protections for minors and users experiencing mental-health crises. For users known to be minors, operators must disclose that the user is interacting with AI, provide break reminders and repeated disclosures at least every three hours during continued interactions, and take reasonable steps to prevent sexually explicit material or encouragement of sexually explicit conduct. For all users, operators must maintain and publish protocols to prevent suicidal ideation, suicide, or self-harm content, including crisis referrals such as suicide hotlines or text lines when needed. Beginning July 1, 2027, operators must annually report certain non-identifying information to the Office of Suicide Prevention, which must post the data online.
The bill’s impact is to add a new layer of consumer-protection and safety regulation to AI companion services in California, with enforcement through private civil actions. A person injured by noncompliance may seek injunctive relief, damages of at least $1,000 per violation or actual damages, and attorneys’ fees and costs. The chapter is cumulative to other laws, meaning it does not replace existing obligations under other statutes, and it applies to operators making companion chatbot platforms available in the state.
Overall, the bill appears to have broad legislative support, with multiple committee and floor votes passing overwhelmingly and only a small number of dissenting votes. The vote history suggests the measure was generally viewed as a safety and transparency bill rather than a controversial expansion of AI regulation. Its focus on minors, suicide prevention, and disclosure likely contributed to the favorable reception.
The main points of contention are likely to have centered on the scope of regulation, compliance burdens for AI companies, and the private right of action. The bill requires operators to publish safety protocols, use evidence-based methods for measuring suicidal ideation, and face civil liability for violations, which could raise concerns among platform operators about implementation costs, litigation exposure, and how broadly the law reaches conversational AI products. At the same time, the bill’s exclusions for customer-service tools, certain video-game bots, and standard virtual assistants indicate an effort to narrow the law to emotionally interactive companion systems.
SB 243 adds Chapter 22.6 to Division 8 of the Business and Professions Code, creating new statutory duties for operators of companion chatbot platforms in California. It requires disclosures that the chatbot is AI, mandates safety protocols for suicidal ideation and self-harm, adds special protections for minors, requires annual reporting to the Office of Suicide Prevention starting in 2027, and authorizes private civil enforcement. The bill also clarifies that these duties are cumulative and do not displace other legal obligations.
The bill appears to have been broadly supported and framed as a consumer-safety and youth-protection measure. Committee and floor votes were largely lopsided in favor, with only a few dissenting votes at later stages. The overall sentiment from the legislative record suggests strong agreement on the need for transparency and safeguards around emotionally interactive AI chatbots.
The likely areas of disagreement were the regulatory burden on AI platform operators, the breadth of the disclosure and monitoring requirements, and the private right of action with statutory damages and attorneys’ fees. Critics may also have been concerned about how operators would reliably detect suicidal ideation, how often minors would receive reminders, and whether the law could be difficult to apply to rapidly evolving AI products. Supporters, by contrast, appear to have prioritized preventing deception, self-harm, and sexual exploitation of minors.