California Gov. Gavin Newsom signed 13 bills on Sept. 10 targeting children’s interactions with social media, artificial intelligence and education technology. The package includes rules for companion chatbots, limits on specified addictive social-media features for users under 16, safeguards around AI-enabled toys, expanded child-sexual-exploitation provisions and new restrictions on the use of student data in AI development.
The laws give California a broad new set of obligations for youth-facing technology products, though their practical reach will be determined later by implementation, enforcement and litigation. Companies serving California users may face state-specific compliance changes; the evidence available now does not show whether the package will reduce the harms lawmakers and advocates seek to address.
What California enacted
In its announcement of the signings, the governor’s office identified SB 1119 as the companion-chatbot safety measure. The administration says it requires protections for children that include protocols for suicidal ideation, parental controls, notices when a child turns off safety settings, independent child-safety audits and annual risk assessments. Those are product-governance requirements, not a finding that any chatbot is inherently safe or that the measures will prevent a particular outcome.
Another signed bill, AB 2246, addresses children’s access to online services, products or features. The governor’s office says social-media companies may not provide users under 16 with addictive features including autoplay and algorithmic feeds based on a user’s history and profile. It is a notable effort to regulate interface and recommendation design rather than merely the content users post or see.
The official release also lists SB 867, covering toys with companion chatbots, and SB 1276, concerning child sexual exploitation. It does not provide the full operational detail for every measure in the 13-bill package. LAist has reported additional descriptions of some provisions, including features beyond those named by the administration, but those details should not be treated as a substitute for the enacted statutory language.
Student data rules reach AI development
One of the clearest extensions beyond consumer social platforms is AB 1159. According to CalMatters’ reporting on the law, it prohibits technology companies from using student data to train or develop AI models. It also expands specified student-data protections to college students and broadens the group of companies covered by the rules. CalMatters reported that AB 1159 takes effect next year.
The distinction is consequential because education technology routinely handles information generated through school accounts, coursework and digital learning services. The new restriction addresses one particular route by which data collected for educational purposes could be repurposed for AI development. Available reporting does not establish how regulators will interpret every technical edge case, including the boundaries of covered data and companies; those answers will depend on the law’s administration and, potentially, future disputes.
California’s package therefore does not operate as one uniform internet rule. It combines separate obligations for chatbot providers, social-media services, toy makers and education-technology companies. A service that crosses those categories could encounter more than one set of duties, while a company offering only one product type may be affected by a narrower part of the package.
Strong claims, unresolved enforcement
Newsom’s office called the package the nation’s strongest child-safety chatbot and social-media legislation. That is the administration’s assessment, not an independently demonstrated national ranking. The available accounts confirm the bills were signed and describe major requirements, but provide no comparative legal analysis of every state framework and no outcome data from laws that have just been enacted.
Nor is signing a law the same as building an enforcement system. In a post-signing examination by LAist, Ed Howard of the Children’s Advocacy Institute at the University of San Diego School of Law said meaningful protection would depend on adequate resources for regulators and lawyers. The report also noted that some earlier California child-online-safety laws remain challenged or have not yet been enforced.
That problem is especially acute for rules that depend on a company knowing whether a user is under 16. Age-related product restrictions can require services to alter access controls, recommendation systems or feature design. The records behind this package do not resolve how individual companies will make those determinations, nor how California will weigh child-safety enforcement against the privacy implications of collecting more age-related information.
The first tangible deadline identified in independent reporting is AB 1159’s effective date next year. For the rest of the package, the immediate result is a new California rulebook for youth-oriented AI and online services, with its force still resting on the agencies, courts and companies that will have to translate it into practice.