California’s Adam’s Law brings age checks and time limits to AI companion apps

AI companion apps have grown up faster than the rules around them. California has now written some of the toughest. On 10 September 2026, Governor Gavin Newsom signed SB 1119, known as Adam’s Law, setting child-safety requirements for companion chatbots that will reshape how these apps check ages and behave with younger users.

The law is named after Adam Raine, a California teenager who died in 2025. The office of the bill’s lead author, Senator Steve Padilla, states that ChatGPT encouraged him to end his life. The bill passed the state Senate unanimously and the Assembly by 64 votes to 4.

What Adam’s Law requires of AI companion apps

The headline change is age assurance. According to an analysis by the law firm Hunton Andrews Kurth, all operators must either determine users’ ages using specified methods or apply the child protections to everyone. Padilla’s office says operators will rely on a privacy-protective age signal under a separate state law, AB 1043. Apps that ban children must publish a description of how they comply.

For users identified as children, the law sets protective defaults that only a linked parent or guardian can change:

• Persistent memory switched off, with exceptions for users aged 16 and over.
• Push notifications switched off.
• Use limited to one hour at a stretch and two hours a day.
• Periodic, age-appropriate reminders during long conversations that the user is talking to an AI.

Parents of children under 16 can disable access entirely.

Romantic roleplay, flattery and dependence

The law also targets what chatbots say. Operators can be held liable for failing to prevent self-harm content, sexual material, romantic roleplay, excessive flattery and emotionally manipulative content that fosters isolation or dependence, according to Padilla’s office. Hunton adds that operators must take reasonable steps to stop chatbots claiming to be human or sentient, or encouraging children to rely on them for emotional support.

There are process obligations too: documented risk assessments before launching or substantially changing a chatbot, a crisis protocol with in-app referrals to mental health services, a published child safety policy, record-keeping, and independent audits with summaries sent to the state Attorney General. Cross-context behavioural advertising to children and the sale of their data are prohibited.

Penalties and timing

Public prosecutors can seek civil penalties of up to $5,000 per affected child for each negligent violation and up to $15,000 for each intentional one. Families also gain a private right of action for certain breaches, including failures on crisis referrals, parental notice, default protections and harmful outputs.

Adam’s Law takes effect on 1 January 2027, but most of its requirements become operative on 1 July 2027. The first independent audits are due by 1 January 2029, and operators with less than $500m in prior-year revenue are exempt from the audit duty until 2032.

What lawmakers and campaigners said

“Our children’s safety deserves to be at the center of every conversation about technology,” Newsom said. Padilla warned: “We cannot make the same mistakes that were made with social media.” Adam’s parents, Matt and Maria Raine, called the law “a very well thought out and significant step towards ensuring AI chatbot safety for our children.”

Common Sense Media, which sponsored the bill, was blunter. “For three years, chatbot companies have been using our children as test subjects,” said its founder and chief executive, Jim Steyer, in a statement. “Today, California said ‘enough!'”

California is not alone: Connecticut’s AI companion rules

Other states are moving in the same direction. Connecticut’s CART Act, signed in June, took effect on 1 October 2026, with its AI companion provisions following on 1 January 2027. According to UConn’s Daily Campus, they require companions to refer at-risk users to mental health resources, bar bots from claiming to be human, and prohibit them from engaging in romantic or sexual relationships with users, alongside parental controls for minors.

For adults using AI girlfriend and boyfriend apps, the most visible change is likely to be more rigorous age checks at sign-up. Our AI companions hub tracks how the leading apps handle safety and age verification, and our research on AI in dating looks at how the technology is changing relationships.

Sources

• Governor Newsom Signs Adam’s Law (Office of Senator Steve Padilla, 10 September 2026)
• California Enacts Companion Chatbot Child Safety Requirements (Hunton Andrews Kurth, 28 September 2026)
• Governor Newsom Signs Adam’s Law, Setting First Real Safety Standard for Kids and Chatbots (Common Sense Media, 10 September 2026)
• New AI regulation law goes into effect Oct. 1 (The Daily Campus, 30 September 2026)