Google begins labeling AI-generated ads but leaves third-party tools on the honor system
A new disclosure panel spans Search, YouTube, and Discover. Whether the labels are complete depends on what advertisers choose to report.
Tap the My Ad Center panel on a Google ad and you may now find a section titled How this ad was made, flagging whether the ad was created or altered with artificial intelligence. Google has begun rolling the labels out across Search, YouTube, and Discover, and presents the system as a consumer transparency measure. For the studios and video AI creators supplying generative content to advertisers, it introduces a new layer of accountability, applied unevenly.
##Two pathways, one of them unverifiable
The label reaches users through two different routes. When an advertiser builds or modifies assets with Google's own generative AI tools, the disclosure is applied automatically, with no human decision involved. When the advertiser uses third-party tools, the category covering most independent video generation, image synthesis, and voice cloning platforms, the label appears only if the advertiser reports the AI use themselves.
Coverage of the rollout, including TechRepublic's, describes that second pathway as an honor system, and critics quoted in that reporting warn it could leave users with a false sense of how complete the labeling is. An advertiser who declines to disclose third-party AI involvement faces no automated check. As generative video becomes harder to distinguish from live-action production, the practical effect is that a missing label tells users nothing. It could reflect the absence of AI or simply the absence of disclosure.
Get the latest model rankings, product launches, and evaluation insights delivered to your inbox.
##Regulation is setting the pace
Reporting on the rollout ties the policy to regulatory mandates advancing in the EU, India, and New York, with the EU AI Act's phased implementation including provisions on synthetic imagery in commercial contexts. Industry commentary aimed at e-commerce advertisers has read the move as Google aligning with current and upcoming legal obligations rather than acting on internal initiative alone.
Transparency programs built ahead of regulation tend to be designed with compliance optics in mind, and this one fits the pattern. The automatic labeling of Google's own tools can be verified. Self-reporting for everything else cannot be verified at scale, and the design accepts that. The jurisdictional patchwork compounds the problem for brands advertising internationally, since what counts as AI-generated or digitally altered under New York's rules may differ from the EU's definitions. Google's system does not resolve those definitional questions. It hands them to the advertiser.
##What changes for production workflows
Search Engine Journal reports that advertising teams are already adding AI disclosure to their pre-launch review processes. Studios and freelance creators who deliver finished assets to agencies should expect disclosure status to become a standard deliverable alongside the work itself, including documentation of which tools were used and whether the output crosses the threshold for labeling.
The threshold itself is the open question. In a hybrid workflow, where a human editor uses an AI tool to rotoscope a single element of an otherwise live-action spot, it is unclear where Google draws the line, and the available coverage does not specify. The policy targets content created or edited using AI, but that granularity will be tested as real campaigns move through the system.
Even an imperfect label from a platform of Google's scale sets a baseline expectation that synthetic content gets identified, and expectations of that kind tend to migrate into contracts and other platforms' policies. The nearer test is regulatory. As the EU AI Act's synthetic-content provisions phase in, Google's honor-system pathway will either satisfy regulators or require a second draft.