On October 13, 2025, California Governor Gavin Newsom signed SB 243, establishing new safety requirements for AI companion chatbots. As conversational AI systems increasingly simulate ongoing human-like interactions and relationships, the law places new responsibilities on operators to protect users, including minors.

For minors, operators must clearly disclose that the interaction is artificially generated, provide periodic reminders to take breaks, and prevent companion chatbots from generating sexually explicit images. The law also requires protocols for responding to expressions related to suicide or self-harm and reporting certain information to state authorities.

Comment from GAME-F

The challenge of AI is moving beyond “what does AI say?” toward a deeper question: “to what extent should AI substitute for human relationships?” California’s approach is notable because it does not simply prohibit AI. Instead, it requires safeguards based on the possibility that children may perceive AI as human-like or engage with it for extended periods.

At the same time, disclosures and break reminders alone may not be sufficient to maintain an appropriate psychological distance from AI. For children, is AI a useful tool, a source of advice, or something resembling a friend? Society increasingly needs to consider where those boundaries should lie.

References

California Governor — Child online safety legislation, October 13, 2025

California Legislature — SB 243 Companion chatbots