Google Ads extends AI labels to California, Korea and Vietnam

The overlay mostly depends on you. Mark an asset as AI-made or AI-edited in the label setting and a campaign aimed at one of those places shows it. Google may also label assets itself in some cases, such as where it is legally required to, and those labels cannot be overwritten. Its Merchant Center product data specification now cites AI regulations in the same six regions as the reason certain AI-made or AI-edited ads need a disclosure or a label.

Google keeps its distance from the legal result. The setting carries no promise that you meet any specific law, the help page says, and it sends you off to get legal guidance. A label you add in your own design tools is allowed too: Google says it counts as neither a banned text overlay nor a banned watermark.

The California law we track comes later: our disclosure guide, updated on 6 November, gives 1 January 2027 for SB 1050, which covers a prominent synthetic performer, voice included. Our Meta guide, last checked 30 October, lists Meta’s self-disclosure regions as the European Region, California, New York, India and Taiwan.

Set the two lists side by side and they split on three places. An AI asset running on both platforms gets the regional treatment on one only: Google’s overlay in South Korea and Vietnam, Meta’s self-disclosure option in Taiwan.