Google has rolled out a new disclosure feature for ads on Search, YouTube, and Google Discover. Users anywhere in the world can now tap the three-dot or info icon on an ad and see a section called "How this ad was made," which will show whether AI was involved in creating or editing it.
The system is automatic only when the ad was made using Google's own AI advertising tools. For everything else, the advertiser has to manually flag it. Google has confirmed it will not independently verify whether an ad made with outside tools is properly labeled.
This is where the announcement starts to look less solid. The vast majority of AI-generated ad content is produced with third-party tools: image generators, background-swap software, AI copywriting platforms, and so on. Under Google's current setup, those ads get a label only if the advertiser voluntarily adds one.
Meta has been running a similar "AI info" label on its platforms, and its approach is meaningfully stricter. Meta scans ads for industry-standard technical markers embedded by AI tools and applies the label automatically when it detects them. It is not a perfect system either, since those markers can get stripped during export, but it is a more active effort than simply asking advertisers to self-report.
The broader picture matters here. Research by the advertising industry's trade body found that nearly a third of all connected-TV commercials, social media videos, and online videos now use AI tools in production, a number expected to rise further into 2026. Advertisers are using AI partly because it is cheaper: product photography done with AI costs a fraction of a real shoot, and copy can be generated in seconds.
Consumer sentiment is more complicated than the industry assumes. Only a small fraction of consumers say they fully trust ads created entirely by AI. Nearly half say disclosure is essential to building any trust at all. Younger consumers in particular have grown more skeptical of AI-generated ads even as advertiser use of these tools has accelerated.
The legal environment is tightening from multiple directions. In Europe, the EU AI Act's transparency rules take full effect in August 2026, requiring that AI-generated content in advertising be disclosed, with penalties reaching 15 million euros or 3% of a company's global annual revenue, whichever is higher. These rules apply to any company whose ads reach European consumers, regardless of where the company is based. In the US, New York's AI Disclosure Law came into force in June 2026, requiring clear labels on ads where AI produced a substantial portion of the content.
For businesses that run advertising, there are two practical things to take from this. First, if you use any AI tool to produce ad images, copy, or video, you now have a formal mechanism on Google to disclose that, and you should use it: the legal and reputational direction of travel is clearly toward more disclosure, not less. Second, if you buy media or manage agency relationships, it is worth asking your agency whether their creative workflows include tracking which assets were AI-generated, because the compliance responsibility ultimately sits with the advertiser, not the platform.
Google's move is a step in the right direction, but the version it has shipped places most of the burden on advertisers to be honest about what their tools did. The label will be most reliable on ads that Google's own tools produced. For everything else, trust the disclosure only as much as you trust the advertiser.