Google is rolling out a new feature meant to help people understand when an ad they’re seeing on Search, YouTube, or Discover was made using AI. Until now, that kind of disclosure was only required for election ads, even though generative tools have made it just as easy to fabricate a product photo, a testimonial, or a brand scene that never existed anywhere in the real world. Extending a rule built for political ads into the much larger world of commercial advertising is what NewsTrackerToday turns on as the actual scope of this change.
The mechanism sits inside Google’s existing “My Ad Center” panel, accessible through the three-dot menu or info icon on any ad a person encounters. That panel already lets users block an ad, report it, or learn why they were shown it in the first place. Now it adds one more option: “how this ad was made,” which indicates whether AI played a role in creating or editing what the viewer is looking at.
Isabella Moretti reads the enforcement gap built into the design: “When an advertiser uses Google’s own generative AI ad tools, the disclosure switches on automatically, no extra step required. But if the ad was made somewhere else entirely, using any other AI tool, the advertiser has to manually flag it themselves. Google isn’t running its own detection check to catch ads that should be labeled but aren’t. That’s a meaningful gap between a policy that sounds comprehensive and an enforcement mechanism that only really works for content made inside Google’s own tools.” That gap between an automatic check and a self-reported one is what NewsTrackerToday digs into as the actual mechanism behind this week’s announcement.
That gap matters more than it might first appear, given how much AI-generated advertising content now gets produced entirely outside any single platform’s own creative tools, then simply uploaded for distribution. An advertiser using a third-party image generator to build a product scene has every incentive to skip the optional disclosure step, and nothing in Google’s current design catches that omission after the fact.
Sophie Leclerc, who covers the technology sector, lands with a comparison to how detection tools have handled similar gaps elsewhere: “This is the same self-reporting weakness that shows up in deepfake-detection systems generally. A watermarking or labeling system only works reliably when the tool that made the content also participates in flagging it. Anything made outside that ecosystem passes through with no signal attached at all. Google’s ad-disclosure feature has exactly that same structural limitation, just applied to commercial ads instead of viral hoaxes.” That structural gap, more than the feature’s existence, is what NewsTrackerToday circles around as the more durable story here.
In some markets, an ad may still get labeled as AI-made regardless of advertiser cooperation, if local law requires the disclosure outright rather than leaving it optional. That patchwork means the actual coverage of this feature will vary meaningfully by country, even though Google is framing the rollout as a single global change.
Whether advertisers who skip the disclosure step face any real consequence beyond a missing label, or whether regulators in markets with mandatory AI-ad rules start actively enforcing against the platforms that host undisclosed content, is what News Tracker Today wraps in as the open question this rollout leaves for regulators to answer next.
