California AI Child-Safety Initiative Enters Circulation

What changed

The Secretary of State announced that a proposed initiative statute on child safety requirements for AI products, including chatbots, entered circulation. The Attorney General's official title and summary says it would require AI providers to use age-estimation technology, assess and publish child-safety risks and mitigations annually, create parental controls, take reasonable steps to prevent harmful, risky, or manipulative AI content for minors, undergo annual independent safety audits, and report findings to the Attorney General.

• Proposed ballot initiative in circulation, not enacted law.

Who is affected

AI providers only if the proposal qualifies for the ballot, is approved by voters, and later takes effect. No current obligation is created.

My read: the work now is to understand the direction, map exposure, and prepare before the text hardens.

Practical consequence for the US professional

Map the workflow that touches this rule or guidance, decide who owns the sign-off, and keep the primary source handy. If this is a draft or consultation, the work is to prepare and respond. If it is enacted or signed, the work is to adjust operations and records to match the legal status the spec states.

Cross-border duty hook

For a US team, the trigger is not where the office sits. It is whether your people, systems, or vendors touch the jurisdiction or data flow named in the spec. If they do, treat the cited rule, statute, or guidance as part of the operating environment and verify the primary source before you rely on it.