OpenAI has publicly backed California Senate Bill 1119, a youth AI safety proposal that the company says would establish stronger safeguards for products used by minors. In an August 31 policy post, OpenAI urged Governor Gavin Newsom to sign the bill, framing it as a state-level response while federal AI rules remain unsettled.
That support should not be confused with the bill already being law. OpenAI is describing the legislation it wants enacted and the safeguards it believes the measure would require. Its post highlights age determination, safety-risk work before release, independent audits, parental tools, crisis-support pathways, limits around harmful content and restrictions involving targeted advertising and privacy.
What OpenAI says the bill would require
The company's case for SB 1119 is built around making youth protections a product obligation rather than an optional setting. OpenAI says covered AI services should determine or estimate age so younger users can receive appropriate safeguards. It also points to pre-release safety assessments and independent auditing as ways to test whether those protections work before and after products reach users.
OpenAI's summary emphasizes controls for families as well. It highlights parental tools, pathways that can connect young people with crisis support, and stronger restrictions on content that could be harmful to minors. The company also says the proposal addresses targeted advertising and privacy for younger users.
Those are OpenAI's descriptions of the bill and its policy position, not an independent legal interpretation. The practical requirements will depend on the final enacted text and any implementation rules if Governor Newsom signs it.
How this fits OpenAI's teen product changes
The advocacy follows OpenAI's August launch of ChatGPT for Teens. The company says accounts it estimates to belong to someone under 18, or accounts where the user states they are 13 to 17, are automatically placed into that experience. OpenAI describes it as the same core ChatGPT product with stronger age-appropriate safeguards, plus family tools and prompts designed to encourage balanced use.
OpenAI also says linked parents can manage selected settings and receive limited safety notifications in certain high-risk situations. Its product documentation says parental controls do not let a parent read or monitor a teen's conversations. That distinction matters because youth-safety tools can otherwise slide into broad surveillance claims that the product does not actually make.
By supporting SB 1119, OpenAI is arguing that parts of this safety model should be reflected in rules that apply more broadly than one company's own product design. There is an obvious policy interest in that position: statewide requirements can set a common baseline for competitors as well as OpenAI.
The next concrete step is political rather than technical. OpenAI has asked Newsom to sign SB 1119. Until that happens, the accurate description is that OpenAI backs the bill and is advocating for its enactment; it is not a settled California requirement yet.
That distinction will matter as the bill moves through the final state process. Product announcements can be changed by a company; legislation can impose duties across covered services. OpenAI is supporting that broader baseline, but the obligations are not final until the legislative and gubernatorial process is complete.
Reporting notes