California bans addictive social media features for children under 16 and tightens AI chatbot rules

California has implemented new regulations to curb addictive social media features and tighten rules around AI chatbots for children under the age of 16, with Governor Gavin Newsom signing a package of laws on September 10. The legislation, described as the strongest in the country, includes several key provisions. Firstly, social media companies are barred from offering addictive features like autoplay and algorithmic feeds to users under 16, which can include personalized content based on user history and profile information.

Secondly, SB 1119, or Adam's Law, named after a teenager who allegedly committed suicide after using ChatGPT, mandates risk assessments and independent audits for new or substantially modified companion chatbots. Operators must utilize age-bracket signals provided by operating systems and impose default protections such as usage limits, disabled notifications, and restrictions on persistent conversational memory.

Failure to address harmful outputs like sexual content or emotional manipulation can lead to liability. The law also requires incident reporting and submission of audit findings to the California Attorney General. Additionally, the package restricts targeted advertising to children, limits the use of school pupil data in AI systems, and allows families to opt out of school-issued laptops.

These measures aim to prevent a repeat of past social media issues and prioritize child safety. Companies, including Meta, have expressed concerns over the restrictions, arguing that personalization aids in providing age-appropriate content.