California will require disclosure on video and audio ads that prominently feature AI-generated people starting Jan. 1, 2027, under a law Gov. Gavin Newsom signed Sept. 16.
This lands on any Ecommerce Paradise reader who runs ads with an AI avatar presenter, a synthetic voiceover or a computer-generated “customer” reacting to the product. Those ads need a clear label before the new year, and the people who can enforce the rule include state prosecutors and shoppers who say they lost money.
Three fresh data points sharpen the timing. Resemble AI’s ad report, covered by PPC Land on Oct. 27, found that only 2 of 3,379 AI-flagged Meta ad files mentioned AI in the ad copy. Etsy tightened its AI image rules on Oct. 22. New York’s version of the law has been in force since June 9. What follows is what California’s law says, what the data shows, and a five-step audit you can finish this month.
An ad that breaks a state advertising law lands on your company, so form your LLC with Northwest, the best LLC formation service for ecommerce owners. Northwest has run for about 25 years under steady ownership, which matters more than a low first-year price from a formation mill that may not be around at renewal. Start your LLC the right way with Northwest →
California SB 1050 Requires Labels on AI Performers in Ads
Newsom signed SB 1050 on Sept. 16, and it was chaptered the same day as Chapter 246 of the Statutes of 2026, according to a Recording Law analysis published Sept. 18. The bill has no urgency clause, so it becomes operative Jan. 1, 2027. It adds Section 17610 to the state’s Business and Professions Code.
The statute covers any “advertisement,” defined as an audio, video or audiovisual communication meant to induce a purchase, per a Vorys client alert dated Oct. 1. A “synthetic performer” is a figure, voice or representation made wholly or partly with generative AI that gives the impression of a human performer who is not recognizable as an identifiable real person. The disclosure duty applies when that performer is included prominently.
Vorys lists three ways an ad crosses the “prominent” line: the performer demonstrates the product in the foreground, provides or voices the narration, or reacts to the commercial message. The label must be “difficult to miss, easily understandable,” and noticeable to a reasonable consumer. An Ecommerce Alliance summary dated Oct. 5 suggests wording such as “This performance features a synthetic performer” or “No human performer is depicted.”
Two exemptions exist. Ads for expressive works such as films, TV shows and video games are exempt when the performer is used consistently with the work, and so is AI used solely for language translation or accessibility. The Ecommerce Alliance summary adds that AI used for ad copy, ideas or other behind-the-scenes work is not covered by itself. The sources describe the law as reaching video, audio and audiovisual ads, which leaves still images outside it on their face.
Enforcement runs through California’s false advertising and unfair competition statutes. Per Vorys, the Attorney General, district attorneys, certain city and county attorneys and injured persons who lost money or property can bring claims. None of the three alerts states a fine amount.
Platforms carry a lighter burden. Recording Law reports that ad mediums must stop distributing a flagged ad only after a court order identifying it is served on them, with no duty to screen on their own. That puts the responsibility on the advertiser, not on Meta, Google or TikTok.
The Ecommerce Alliance summary says out-of-state businesses may be affected if their ads reach California consumers. Vorys titled its alert with a prediction that the rule “is sure to lead to wave of lawsuits.” That is a law firm’s forecast, not a fact about what will happen.
New York’s Penalties, Etsy’s Photo Rule and Meta’s Label Gap
California is the second state to act. New York Gov. Kathy Hochul signed S.8420-A/A.8887-B on Dec. 11, 2025, and it took effect June 9, 2026, according to Cooley. The New York civil penalty is $1,000 for a first violation and $5,000 for each later one. New York exempts audio-only ads and the platforms that carry ads, while California covers synthetic voices in audio advertising, per the Ecommerce Alliance summary.
The counterpoint matters. Cooley noted that New York’s law may face federal opposition and First Amendment challenges, citing a Dec. 11, 2025 executive order on AI regulation. The same arguments could reach California’s statute. None of the alerts I read reports a court ruling on either law, so plan for the rule to be live on Jan. 1.
Marketplaces are moving on their own clock. Etsy published new rules on Oct. 22, and Digital Commerce 360 reported on Oct. 26 that every listing now needs at least one unedited photo of the actual item on a real background. AI-edited supplemental images stay allowed if they accurately represent what the buyer receives. Etsy also banned AI-generated patterns, recipes and instructions and said bot-generated listings will be removed.
Retail Dive added on Oct. 23 that sellers may have to refund buyers who open a case claiming the product differs materially from the listing, if the listing failed the new image rules. James Ossman, Etsy’s vice president of customer operations, said, “Buyers should always be able to understand the role you play in every item you sell on Etsy.” PPC Land reported that Etsy’s post gives no effective date and no transition period. Etsy is a marketplace policy, not a law, and most high-ticket Shopify stores do not sell there. It shows where platform enforcement is heading.
The Meta data shows the gap between policy and practice. Resemble AI published its 2026 Synthetic Advertising Report on Oct. 22, and PPC Land covered it on Oct. 27. Resemble scanned 37,768 image and video files from Facebook and Instagram ads and flagged 3,379 of them, or 8.9%, as AI-generated or manipulated.
Brand pages ran at a lower rate: 1,971 of 28,463 files, or 6.9%, and 222 of 389 brands ran at least one flagged file. Consumer packaged goods led industries at 13.4%. Only 2 of the 3,379 flagged files mentioned AI in the ad copy, and none of the 971 files checked for C2PA provenance metadata carried it. Meta announced a “Made with AI” label in April 2024 and signed the EU’s voluntary transparency code in July 2026, per the same report. Meta did not comment.
Read the report with care. Resemble sells synthetic media detection software, every verdict comes from its own detector, and it published no error rates. It also measured written disclosure in ad copy, not on-screen labels. One vendor’s scan is not proof that 8.9% of ads are fake, but it does show how many accounts are already running AI visuals with no disclosure.
Platform rules are piling up on advertisers. I covered Meta’s terms that make you liable for its AI ad edits starting Oct. 30. I also flagged that ChatGPT ads can rewrite your copy and URLs unless you opt out. California adds a legal layer on top of both.
Which Store Ads Trigger SB 1050 and What a Miss Could Cost
My read is that the law is narrower than the headlines suggest and sharper than most store owners expect. It only bites on realistic AI humans in video or audio ads. Most high-ticket stores have a few of those, and a few is enough.
The most exposed assets are the ones I have seen recommended everywhere this year. An AI avatar fronting a product page explainer, which I walked through in my guide to AI avatar explainers for high-ticket product pages, meets two of Vorys’s three triggers: the avatar demonstrates and voices the pitch. A fake “customer” testimonial built the way my breakdown of UGC-style ads without filming describes meets the reaction trigger.
A synthetic voiceover on stock footage is the sneaky one. The Ecommerce Alliance summary says California reaches synthetic voices, and a voice that narrates the ad is the “provides the narration” trigger. If you generate presenter videos in InVideo, check whether the output includes a synthetic voice or face before you upload it.
What stays outside the law on its face: AI-written ad copy, AI-generated product shots with no person in them, and AI used only to translate a real human’s speech. My broader view on still images is in my comparison of AI-generated and real product images, and Etsy’s new photo rule backs the same conclusion: real photos of the real item win on trust.
Product page videos are the gray zone. The statute defines an ad as a communication meant to induce a purchase, and a video on a Shopify product page is arguably that. No source I read addresses it, so treat it as covered until counsel says otherwise. I’m not a lawyer, this is reported information, not legal advice, and a service like LegalShield is one way to get an attorney to review your ad library.
Meta’s generative tools add a second risk. Under the Oct. 30 terms, advertisers carry the risk for Meta’s AI edits. If an Advantage+ creative feature inserts a synthetic person into your ad, you could be the party that “included” one. That is my interpretation, not a ruling, and the safe move is to review which generative enhancements are switched on in your ad account.
Here is hypothetical math, labeled as such. Say a store runs 40 active video ads and 12 use an AI presenter. New York’s penalty structure is $1,000 for a first violation and $5,000 for each later one. If regulators counted each ad as a separate violation, that is $1,000 plus 11 times $5,000, or $56,000. How violations are counted is untested, and California’s alerts state no fine at all. The real cost in California is more likely a private suit or an Attorney General demand letter, plus the cost of pulling ads mid-campaign.
The decision comes down to thresholds. With zero AI people in your video and audio ads, you only need contract language that keeps it that way. With one to three, label them and move on. With a library built around AI presenters, you face a real choice: label everything and test the conversion hit, or replace the presenters with real creators from a marketplace like Fiverr. I have no data on how a “synthetic performer” label changes click-through or conversion rates, so test it on a small budget before you roll it out.
High-ticket stores have more to lose here than low-ticket ones. A buyer spending $2,000 or more on a product usually checks who stands behind the claim, and a fake customer reacting to your product is the kind of thing that turns into a dispute. My chargeback prevention guide for high-ticket stores covers why a misled buyer is a chargeback waiting to happen. The same trust logic runs through my pillar guide on what high-ticket dropshipping is.
If you would rather hand creative production and compliance to a team that builds this in from day one, that is what the done-for-you turnkey build is for.
Want 1-on-1 coaching to audit your ads and fix your high-ticket store before Jan. 1? Get the coaching details →
Five Steps to Audit Your Video and Voice Ads Before Jan. 1
You have 65 days. Here are five steps, in order:
- Inventory every video and audio asset. List ads in Meta Ads Manager, Google Ads, YouTube and TikTok, plus product page videos and any video inside Klaviyo email flows. Hand the spreadsheet job to a virtual assistant from OnlineJobs.ph and have them mark every asset with an AI face or AI voice.
- Sort each flagged asset into label, replace or cut. Replace is usually the cheapest long-term answer for testimonial-style ads, because a real creator removes the question. My guide to AI video generation for ecommerce ads covers where AI video still earns its place, such as product motion with no human in frame.
- Burn the label into the video. Use on-screen text that is large, high contrast and visible for the whole clip, since the statute demands a label that is “difficult to miss.” A template in Canva makes this a two-minute job per asset. Use wording close to the suggested lines above.
- Rewrite your creator and agency contracts. Add a clause that no synthetic performer is used without written notice, and ask Meta-facing agencies which generative enhancements they have enabled. Then turn off any enhancement that adds or alters people unless you are ready to label the output.
- Keep a dated log and update your site policies. Record each ad version, whether it contains a synthetic performer and when the label went live. Pull the disclosure language into your site policies with a generator like Termly. Then have your team run every new creative through a one-page review checklist, which you can draft in minutes with Claude.
Explore Turnkey Builds for service details and pricing. Email me with questions before getting started.
Frequently Asked Questions
Does SB 1050 apply to my store if I am not in California?
It may. The Ecommerce Alliance summary says out-of-state businesses can be affected when their ads reach California consumers. Treat any nationwide video campaign as in scope and confirm with a lawyer.
What counts as a synthetic performer?
Per Vorys, it is a figure, voice or representation made wholly or partly with generative AI that gives the impression of a human performer who is not recognizable as a real, identifiable person. An AI avatar, a synthetic voice or a generated “customer” all fit. See my roundup of AI content platforms for ecommerce to check which tools produce them.
Do AI product photos need a label under the law?
The alerts describe the law as covering audio, video and audiovisual ads, so still images fall outside it on their face. Marketplaces and ad platforms have their own rules, though, and Etsy now wants one unedited photo of the real item in every listing. Read my coverage of eBay’s wider AI monitoring of seller listings if you sell there too.
What should the label say?
The statute asks for wording substantially similar to “This performance features a synthetic performer” or “No human performer is depicted.” Make it readable on a phone screen for the full length of the clip.
How big is the fine?
None of the California alerts states one. New York’s is $1,000 for a first violation and $5,000 for each later one, per Cooley. In California the more likely exposure is a claim under the false advertising and unfair competition statutes.
Will Meta or Google add the label for me?
Do not count on it. Recording Law reports that platforms must act only after a court order and have no duty to screen ads, and Resemble’s data suggests disclosure on Meta is thin today. Check what Meta’s delayed commercial terms require of you as well.
What if a UGC creator used AI without telling me?
The advertiser is the party the law targets, so you carry the risk. Put an AI disclosure warranty in every creator agreement, and grab my free high-ticket niches list if you are still choosing products to advertise.
Want my team to scale the store you already have, with compliant ad creative built in? See the scaling service →
Subscribe to the YouTube channel for daily breakdowns. More breaking news coming through the day.
Related Articles
If this was useful, these go deeper:
- AI-Generated Vs. Real Product Images
- How to Use an AI Avatar Generator for High-Ticket Product Page Explainers
- How to Create UGC Style Ads for Your Store Without Filming Anything
- How to Start a Business in the USA From Canada: High-Ticket Ecommerce
- High-Ticket Niches List: 150+ Best Dropshipping Niches for 2026
Trevor Fenner is an ecommerce entrepreneur and world traveler from Seattle, Washington, and the founder of Ecommerce Paradise, a platform focused on helping entrepreneurs build and scale profitable high-ticket ecommerce and dropshipping businesses. With over a decade of hands-on experience, Trevor specializes in high-ticket dropshipping strategy, niche and product selection, supplier recruiting and onboarding, Google & Bing Shopping ads, ecommerce SEO, and systems-driven automation and scaling. Through Ecommerce Paradise, he provides free education via in-depth guides like How to Start High-Ticket Dropshipping, advanced training through the High-Ticket Dropshipping Masterclass, and fully done-for-you turnkey ecommerce services for entrepreneurs who want a faster, more hands-off path to growth. Trevor is known for emphasizing sustainable, real-world ecommerce models over hype-driven tactics, helping store owners build scalable, sellable, and location-independent brands.
He maintains homes in Seattle and Bali, spends most of the year overseas, and usually returns to Seattle for about a month each summer.
