A California bill that would regulate AI companion chatbots is close to becoming law (techcrunch.com)

California’s SB 243 — a bipartisan bill that would be the first state law to specifically regulate “companion” AI chatbots — has passed the legislature and is awaiting Gov. Newsom’s signature (deadline Oct. 12). If enacted (effective Jan. 1, 2026), the law would require operators of adaptive, human‑like chatbots that meet users’ social needs to block conversations about suicidal ideation, self‑harm and sexually explicit content; send recurring disclaimers (for minors, every three hours) that the user is interacting with an AI and should take a break; and meet annual transparency/reporting obligations (phased in July 1, 2027). The statute creates a private right of action allowing injured parties to seek injunctive relief, attorney fees and damages up to $1,000 per violation. For AI/ML teams and companies such as OpenAI, Character.AI and Replika, SB 243 raises operational and technical requirements: improved content classification and refusal policies, automated crisis‑referral flows, session timers and persistent identity/age gating, plus new logging and reporting pipelines. The bill was narrowed during negotiations (removing earlier bans on “variable reward” engagement mechanics and some tracking mandates), but still signals heightened legal and regulatory risk—alongside parallel FTC and state investigations—and establishes a precedent for state‑level safety rules that could shape model behavior, moderation engineering, and product design for companion agents nationwide.