SB 1000, Sen. Josh Becker’s rewrite of the California AI Transparency Act, was enrolled on August 27 after concurrence cleared 39-0, and now sits on Gov. Gavin Newsom’s desk with an urgency clause that would make it effective the moment he signs.

The bill’s structural move is a single deletion. CAITA, which became operative August 2 as SB 942 (amended by AB 853), currently applies only to generative-AI providers with more than 1,000,000 monthly California users. SB 1000 strikes that threshold. Per Morgan Lewis, scope would extend to any GenAI system publicly accessible in the state.

That’s the whole story in one line: a statute drafted to police a handful of frontier labs becomes, overnight, a statute that reaches every image, video, and audio generator a Californian can open in a browser.

The rest of the amendments to Bus. & Prof. Code §§ 22757.1, 22757.2, and 22757.3 tighten the regime further. SB 1000 scraps the manifest disclosure option, compresses the licensee revocation window from 96 to 72 hours, and swaps the “AI detection tool” for a “disclosure verification tool.” The penalty architecture is unchanged and unforgiving: $5,000 per violation, each day a discrete violation, enforceable by the Attorney General, city attorneys, or county counsel.

The obligations run to the provider of the GenAI system, not the small operator using it downstream. But small teams shipping AI-generated marketing into California should confirm their vendors are ready. Becker’s bill is one thread in California’s AI bill sprint, alongside the synthetic-performer ad disclosure bill, and it rhymes with the EU AI Act’s disclosure regime.

Newsom has 30 days post-adjournment to sign or veto, per the Transparency Coalition. The threshold that made CAITA a big-platform statute exists, for now, only until he decides otherwise.

Sources

Sources