California SB 1050: disclose AI actors in your ads

If you generate the face or the voice in your client's video ad, California just made that a labeling job. Governor Newsom signed SB 1050 on September 16, requiring a clear disclosure on any video or audio ad that features an AI-generated synthetic performer. New York got there first in June. California is the second state, and it is the one with the ad spend — which means synthetic performer disclosure stops being a coastal curiosity and starts being a line in your creative checklist.

What actually happened

SB 1050, authored by Senator Angelique Ashby and sponsored by SAG-AFTRA, defines a synthetic performer as a digital figure, voice, or representation created with generative AI that creates the realistic impression of a human performance by someone not recognizable as any identifiable natural person. In plain terms: the invented spokesperson, not the deepfaked celebrity. The celebrity case was already covered by California's 2024 digital likeness laws.

The obligation is a clear and conspicuous disclosure when an ad prominently features one. The bill text points at wording substantially similar to "this performance features a synthetic performer" or "no human performer is depicted" — notably more prescriptive than New York, which left "conspicuous" undefined and let everyone guess.

Enforcement routes through machinery that already exists: a violation is a violation of Business and Professions Code § 17500, California's false advertising statute, enforceable under the Unfair Competition Law at § 17200. Advertising mediums — broadcasters, cable operators, online platforms, streaming services — must pull or disable an ad once served with a court order. Carve-outs cover expressive works like film, TV and video games, and cases where AI is used solely for translation or accessibility.

One thing to pin down yourself: neither the Governor's release nor the coverage names an effective date. California statutes from a regular session take effect January 1 following enactment absent an urgency clause, which puts this at January 1, 2027. Plan against that date and confirm it against the chaptered text before you brief a client.

Why it matters for your business

Two states now, with different disclosure wording, and the obligation attaches to whoever produces the ad — not the tool that generated the pixels. If you run paid social for a client and the creative includes an AI actor, that is your compliance surface, not Runway's.

The operational answer is boring and cheap: make the label a field in your asset pipeline, not a legal review. Tag any creative containing a synthetic performer at generation time, carry the flag through to the ad platform upload, and render the disclosure in the asset itself rather than the caption — captions get truncated, reposted, and stripped. New York fines start at $1,000 and rise to $5,000 per subsequent violation; California's route through § 17200 is an injunction-and-removal path, which for a live campaign is arguably worse than a fine.

AI creative got cheap in 2025, and 2026 is the year the paperwork caught up. Two states is a trend line, not an outlier. Build the label into the workflow once and the third state costs you nothing.

Sources: Office of Governor Gavin Newsom, California SB 1050 bill text.