Google Ads expands AI disclosure labels to California, South Korea and Vietnam

Google has added California, South Korea and Vietnam to the list of regions where ads built from AI-generated or AI-edited assets must carry a visible, on-ad disclosure label, according to Google Ads and Merchant Center help documentation updated this week and first reported by Search Engine Roundtable on November 5. The change extends a rule that, until now, only applied to campaigns targeting the European Union, India and New York.

For software marketers, the practical effect lands squarely on two markets most B2B and SaaS companies can't skip: California, home to a large share of the US buyer base many vendors target first, and a Google Ads footprint that increasingly spans South Korea and Vietnam as APAC budgets grow.

What changed

The mechanism itself isn't new. Google Ads has let advertisers mark individual assets as "created or edited with AI" since the labeling setting first rolled out, and Google surfaces that status globally through the "How this ad was made" panel inside My Ad Center, reachable from the three-dot menu on ads across Search, YouTube and Discover. What's new is where that designation also triggers a visible overlay printed directly on the creative itself, rather than staying tucked inside a disclosure panel most users never open.

Until this update, that on-ad overlay applied only to campaigns targeting the European Union, India or New York. Google's help documentation now lists six regions: "In some geographies, AI labels will also show on the ad itself." California, South Korea and Vietnam join the original three. The same language was added to the Merchant Center's product data specification, meaning Shopping listings built with AI-generated product imagery fall under the same rule.

Two details matter for anyone managing campaigns rather than just reading about them. First, Google can apply the label itself in some cases, including where Google is legally required to or where it receives signals from other platforms, and once applied, these labels cannot be removed or overwritten by the advertiser. Second, Google is explicit that flipping the setting is not a compliance guarantee on its own, and that advertisers should seek their own legal guidance.

Who it affects

Any software company running Search, Display, Shopping, Performance Max, Demand Gen or Display & Video 360 campaigns that touch AI-generated or AI-edited images, video or audio, and that target California, South Korea or Vietnam as part of a broader geo set, not just as standalone campaigns, is now in scope. That includes a lot of ordinary SaaS media buying: AI-upscaled product screenshots, generative background art behind a demo video, or AI voiceover on a YouTube pre-roll.

California's inclusion isn't happening in isolation. The state separately enacted its own disclosure law for ads built around AI-generated performers, which Governor Gavin Newsom signed on September 16, 2026, and which took effect immediately. That statute requires a "clear and conspicuous" disclosure whenever an ad prominently features a synthetic performer demonstrating a product or delivering a commercial message, a narrower, performer-specific rule that sits alongside, not inside, Google's broader asset-labeling setting. The two obligations are separate, but they point the same direction: AI-ad disclosure law is arriving state by state and region by region faster than most marketing teams are tracking it.

Barry Schwartz, who broke the Google documentation change at Search Engine Roundtable, noted the pattern across Google's last two expansions and wrote that "I suspect this list will continue to expand over time." Nothing in Google's own documentation contradicts that read; the help page frames this as an expanding regional list, not a finished one.

What to do differently

The fix isn't complicated, but it requires someone to actually go look, because Google doesn't retroactively re-flag assets you uploaded before the setting existed.

• Pull the "AI label" status column in Asset Library and asset reporting for every active campaign targeting California, South Korea or Vietnam.
• Flag any asset with generative fills, AI voiceover, AI-upscaled imagery or synthetic video and confirm its label status before the next campaign refresh.
• Check Merchant Center product images separately. Shopping feeds are covered by the same update and are easy to overlook if your team treats Ads and Merchant Center as separate workstreams.
• Loop in legal or compliance before assuming the Google setting covers you. Google says explicitly that it doesn't guarantee regulatory compliance on its own.

Agencies running media for multiple software clients have the bigger lift: a single account team may be managing a dozen advertisers across all six labeled regions at once, each with its own mix of AI-touched creative. Building a standing checklist into campaign QA now is cheaper than discovering a labeling gap after a regulator or a platform does.