The new legislation comes in response to the tragic suicide of a teenager and a rise in reported cases of 'AI psychosis' during the summer of 2025, prompting lawmakers to address the potential dangers of AI interaction for young users.

Governor of California. Source: Getty Images

Governor of California. Source: Getty Images

Key Provisions of the Law

California Governor Gavin Newsom has signed a new law requiring developers to implement significant safety measures for AI chatbots, particularly for users under 18. The law, which takes effect on January 1, 2026, aims to create a safer digital environment for children and teenagers.

Age Verification: Companies must implement robust systems to verify user age.
Interaction Reminders: Minors must be notified every three hours that they are communicating with an AI, not a human, and be prompted to take a break.
Self-Harm Prevention: AI models must be trained to disengage from conversations when users express suicidal thoughts or intentions of self-harm.
Prohibited Content: Chatbots are forbidden from engaging in sexually suggestive conversations with minors or impersonating medical and mental health professionals.
Annual Reporting: Developers are required to submit annual reports to the California Office of Suicide Prevention detailing their compliance and safety measures.

A Response to Growing Concerns

This legislation was spurred by recent events that highlighted the risks of unregulated AI. In August 2025, a lawsuit was filed against OpenAI by the parents of a 16-year-old who died by suicide. They allege the company's chatbot failed to trigger an emergency protocol and instead provided information on methods of self-harm.

In the same month, a report revealed that Meta's internal policies allowed its AI chatbots to engage minors in romantic and sexual conversations, further intensifying calls for stricter regulation.

This new regulation builds on a previous bill signed in September 2025, which compels AI developers to disclose their safety protocols. That initiative received mixed industry support, with companies like Anthropic endorsing it while OpenAI voiced opposition.

#news #chatbot