OpenAI has come out in support of SB 1119, a California Senate bill that would impose safety requirements on how minors use AI. Ann O'Leary, who leads global policy at the company, published a letter to Governor Gavin Newsom urging him to sign the bill into law[1]. Age determination, independent audits, parental controls: the mechanisms OpenAI already runs in its own products would become the state's minimum standard.
States Set the Standard While Federal Action Stalls
OpenAI frames its position around the absence of federal legislation, arguing that California has an opportunity to set a strong standard for youth AI safety[1]. The letter commends Governor Newsom along with Senator Steve Padilla, Assemblymembers Buffy Wicks and Rebecca Bauer-Kahan, and the California Legislature for their work on the bill[1].
The company says the move builds on principles for teen users that co-founder and CEO Sam Altman outlined last year[1]. SB 1119 is paired with Assembly Bill 2023, carried by Bauer-Kahan and Wicks, and both measures were amended together in March 2026[2].
The context is how deeply AI chatbots have moved into teenage life. Common Sense Media research cited in the March announcement found that 72 percent of teens have used an AI companion and more than half are regular users. The average user spends 93 minutes a day with AI chatbots, 18 minutes longer than on TikTok[2].
What Operators Would Have to Do
OpenAI explicitly backs 7 requirements: determine a user's age; identify and address safety risks before making a product available to young people; undergo independent audits; protect young people from harmful content including self-harm and sexually exploitative material; give parents tools to guide and limit use; connect young people with crisis-support resources when serious risks arise; and limit targeted advertising while protecting personal information. For users identified as ages 13 through 17, the company says these protections should apply automatically[1].
The March amendments went further still. They included mandatory annual risk assessments, submission of audit findings to the California Attorney General, a ban on selling, sharing, or using a child's personal information, and usage caps for minors of 1 hour per session and 2 hours per day[2]. The bill builds on existing California law under SB 243, with the aim of preventing operators from dodging responsibility after harm occurs[2].
ChatGPT for Teens Already Applies Automatically
Part of why OpenAI is comfortable with the bill is that it has already shipped much of what the bill asks for. ChatGPT for Teens is designed for users under 18, and users are placed into it automatically when the system estimates they are under 18 or when they state their age is between 13 and 17. Those protections are part of the baseline experience rather than optional settings that users can turn off[1].
Learning features come with it. Quizzes, Learning Visualizations, and Study mode, which walks students through problems step by step, were developed with educators and learning experts[1]. OpenAI says nearly 9 in 10 teens who use ChatGPT turn to it in a given week for learning, information, skill-building, or productivity[1].
Over the past year the company has added parental controls, age prediction, its Teen Safety Blueprint, and Under-18 Principles in the Model Spec, the rules that govern how its models are meant to behave. Those rules prohibit romantic engagement, encouragement of emotional dependence, and claims that ChatGPT is human or sentient[1].
Drawing a Line Between AI and Social Media
The point OpenAI stresses most is that SB 1119 does not treat AI as social media[1]. In building the framework, lawmakers worked to preserve educational and safety-critical features, including responsible uses of ChatGPT's memory[1].
Memory looks, at first glance, like a feature that accumulates personal data about minors. OpenAI's argument is that carrying context forward lets teens continue long-term learning projects without starting over, while also helping apply safeguards consistently and recognize concerning patterns[1]. Stripping data retention entirely would weaken the safety machinery too, and that tension is likely to keep resurfacing in regulatory debates.
OpenAI adds that continued study and collaboration during implementation can improve key provisions and keep the framework effective as the technology evolves[1].
Summary
With OpenAI backing SB 1119, the standard-setting process for teen AI use has reached the stage where safeguards vendors built voluntarily get fixed in state law. Age determination with automatic enforcement, independent audits, and routing to crisis support all overlap with what ChatGPT for Teens already does. The bill's fate rests with Governor Newsom, but whatever is settled here has a strong chance of becoming the starting point for debates in other states and countries.
Source [1]: https://openai.com/index/supporting-california-bill-advance-ai-youth-safety
