Urgent.News

What's breaking now, across thousands of outlets.

AI

OpenAI Backs California Bill Requiring Age Checks and Safety Audits for Teen AI

OpenAI is backing California SB 1119, which would require age checks, safety protections, and audits for AI companion chatbots used by minors. The post OpenAI Backs California Bill Requiring Age Checks and Safety Audits for Teen AI appeared first on TechRepublic .

OpenAI has expressed support for a proposed California bill, SB 1119, which aims to impose age checks, safety protections, and auditing requirements for AI chatbots used by minors. The bill would make such measures mandatory rather than voluntary, requiring AI operators to verify users' ages or default to child protections, conduct safety assessments before releasing new or significantly modified companion chatbots, publish child safety policies, and undergo independent audits.

Specific protections would be established against harmful interactions, including self-harm encouragement, simulated romantic interest, claims of human or sentient status, and attempts to foster emotional dependence. For younger users, the bill would include default restrictions on continuous daily use, persistent conversational memory, push notifications, and custom system prompts.

Parents would have the ability to manage or disable access for children under 16. OpenAI argues that teenagers and AI use differ from social platforms, advocating for regulations that protect beneficial applications such as learning, research, and productivity while maintaining appropriate safeguards. The legislation, if signed by Governor Gavin Newsom, could establish a strong standard for youth AI safety beyond California and influence other states' regulations.

Written by urgent.news from TechRepublic's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at techrepublic.com →

More in AI

llm-gemini 0.34

Release: llm-gemini 0.34 New model gemini-3.8-flash for Gemini 3.8 Flash , with low, medium and high thinking levels. #146 Fixed async responses failing to record the resolved model version.

More from Wednesday 2 September →