California AI law for physicians: AB 3030 and AB 489

Two California laws every physician using AI should know

One governs what a patient must be told when AI writes to them about their health. The other governs what an AI tool may imply about who is caring for them.

Disclosing AI-generated patient communications — Health & Saf. Code §1339.75 (AB 3030)

A health facility, clinic, physician’s office or group practice that uses generative AI to write or speak to patients about their clinical information must include a disclaimer that the communication was generated by generative AI, and clear instructions for reaching a human health care provider or other appropriate person (§1339.75(a)).

Where the disclaimer goes depends on the medium: at the beginning of letters and emails; displayed throughout chat-based and video interactions; spoken at the start and end of audio (§1339.75(a)(1)(A)–(D)).

The disclaimer is not required when the AI-generated communication is “read and reviewed by a human licensed or certified health care provider” (§1339.75(b)). Administrative messages — scheduling, billing and other business matters — are outside “patient clinical information” (§1339.75(c)(7)).

A physician’s violation falls under the jurisdiction of the Medical Board of California or the Osteopathic Medical Board, as appropriate (§1339.75(d)(3)).

No AI may pass itself off as a licensed professional — Bus. & Prof. Code §§4999.8–4999.9 (AB 489)

The prohibitions that stop an unlicensed person from using a title or letters that imply a health care licence are enforceable against whoever develops or deploys an AI system that uses those terms in its advertising or functionality (§4999.9(b)).

Using a term, letter or phrase that indicates or implies the care, advice, reports or assessments offered through an AI system are being provided by a licensed natural person is prohibited, and each use is a separate violation (§4999.9(c)–(d)). Enforcement belongs to the relevant licensing board, which may seek an injunction (§4999.9(a)).

What this means in practice

Tell patients when AI wrote to them unless a licensed provider read and reviewed it first; give them a way to reach a person; and make sure no AI output, interface or advertisement speaks as if it were the licensed clinician.

How ChironAI™ OM helps

Every AI output in OM carries a server-built AI-use disclosure citing AB 3030 and AB 489, and first-person licensure phrasing (“I diagnose”, “my medical license”) is rewritten to the decision-support register before you see it. Chiron organizes the records and prepares working drafts; the physician reviews, decides and signs.