California is advancing legislation to establish safety guidelines for teenagers using artificial intelligence. Senate Bill 1119, which recently gained public backing from ChatGPT creator OpenAI, aims to protect minors from harmful chatbot interactions while keeping educational tools accessible.
What Happened
California’s Senate Bill 1119 changes how companion chatbots must treat minor users. Under the bill, AI providers must automatically apply safety features for teenagers aged 13 to 17. According to the Transparency Coalition bill guide, the age verification system relies on secure device-level settings instead of requiring teens to upload passports or other sensitive documents. This transfer protects youth privacy while maintaining compliance.
The legislation also closes the "actual knowledge" loophole. As analyzed by the Angelo Law Firm PLLC, chatbot operators must either verify the age of every user or apply minor-safety protections to all users. Violators face civil penalties of up to $5,000 for negligent violations and $15,000 for intentional violations per child. As we previously reported, OpenAI backed this framework because it separates chatbot regulation from social media laws.
The Bigger Picture
AI tools are widely used by students. Data from the Pew Research Center youth data shows that OpenAI's ChatGPT is the most common tool, used by 59% of U.S. teens. According to an MIT Technology Review youth survey report, 57% of teens use chatbots to search for information, 54% use them for schoolwork, 47% use them for entertainment, and 12% use them for emotional support.
Features like AI memory layers create privacy risks. A Field Guide to AI Memory analysis shows that because standard end-to-end encryption limits how an AI searches and summarizes past logs, providers store student profiles in plaintext on their servers. This leaves contextual histories vulnerable to data breaches.
While groups like Google.org invest in leadership for youth digital safety, regulatory compliance is tightening. The Promise Legal compliance audit guide points out that recent amendments to the Children’s Online Privacy Protection Act (COPPA) mandate separate, verifiable parental consent before disclosing minors' personal data for AI training or advertising. School administrators cannot bypass this requirement on behalf of families, as federal regulators declined to create a school-authorization loophole. Nationally, federal lawmakers are advancing bills like the Kids Online Safety Act (KOSA) to enforce a "duty of care" on platforms, as reported by the International Association of Privacy Professionals report.
What This Means for Families
For parents and educators, SB 1119 shifts the responsibility of protection from families to developers. Instead of relying on children to self-report their age, platforms must automatically activate parental controls, restrict targeted advertising, and block high-risk content like graphic material or self-harm.
AI memory features still require attention. Because these systems retain conversations to personalize future interactions, students can unknowingly store private thoughts, family details, and school assignments on corporate servers indefinitely.
What You Can Do
Parents can protect their children's data by managing chatbot settings. Review and delete your teen's chat history, or disable the memory feature entirely. You can also encourage them to use temporary chats for sensitive questions.
Make sure your children's devices reflect their actual birth dates. Upcoming safety systems will rely on these device-level signals to trigger automatic safety features.
Finally, discuss the limits of AI with your children. Explain that chatbots are not friends, can produce incorrect answers, and store the information typed into them.