Google will start telling users when an ad they see was made with AI, rolling out a new 'How this ad was made' option inside its My Ad Center panel across Search, YouTube, and Google Discover. The disclosure is global and available to any user who taps the three-dot menu or info icon on an ad. It marks the first time Google has extended AI-labeling beyond election ads to the entire advertising surface it controls.
The mechanism is split by origin. Ads created with Google's own generative AI ad tools will have the disclosure switched on automatically. Ads produced outside Google's stack rely on the advertiser to flip a new control declaring that AI was involved. Google confirmed it will not perform its own check to determine whether that self-declaration is accurate.
That is the load-bearing detail. My Ad Center already lets users block ads, report them, or see why an ad was shown and who paid for it. Adding AI provenance is a meaningful expansion of what the panel discloses, but the honor-system design means the label is only as reliable as the advertiser filling out the form.
Key facts
- 01Google is adding a 'How this ad was made' option to the My Ad Center panel across Search, YouTube, and Google Discover.
- 02Ads built with Google's own generative AI ad tools will have the disclosure auto-enabled; ads made elsewhere rely on advertisers to self-declare.
- 03Google will not independently verify whether a non-Google-made ad used AI, leaving accuracy dependent on advertiser honesty.
- 04Until now, AI disclosure was only mandatory for election ads on Google's platforms.
- 05The label will also appear automatically in markets where local law requires AI ad disclosure.
The context for the move is straightforward. Generative AI has made it dramatically cheaper for brands to produce ad creative — synthetic product shots, altered environments, AI-composed models — without commissioning real photography. Google already prohibits misleading and deceptive ads under its policies, but a compliant ad can still use synthetic or digitally altered imagery that a consumer would reasonably assume is a real photo. The new label is meant to close that perception gap.
Google's prior AI-disclosure regime was narrow. Since the last US election cycle, political advertisers on Google and YouTube have been required to disclose synthetic content in election ads. Extending the same principle to commercial advertising broadens the scope from a few thousand political buyers to essentially every business that runs Google ads.
Enforcement, though, is the question the announcement does not answer. Google's generative ad tools are used by a fraction of advertisers. The vast majority of AI-assisted creative — a Midjourney background here, a Firefly product mockup there, an AI-upscaled hero image — is produced in third-party tools and uploaded to Google Ads as finished assets. For those, disclosure depends entirely on an advertiser voluntarily ticking a box that could cost them consumer trust if ticked.
In jurisdictions with statutory AI-disclosure requirements, the label will appear automatically where local law demands it. That backstop matters most in the EU, where the AI Act's transparency provisions for synthetic content are now in force, and in a growing list of US states with their own synthetic-media rules for advertising. Elsewhere, the label is opt-in by the advertiser.
The competitive frame is worth noting. Google is the largest digital-ad platform in the world by revenue, and its choices set de facto standards for the industry. Meta, TikTok, and Amazon have all published AI-disclosure policies of varying strictness, but a self-declared, platform-visible label inside a user-facing dashboard is a step further than most rivals have taken publicly. It also raises the awkward question of whether other platforms will now match — or whether they'll argue that self-declaration without verification is a fig leaf.
The obvious weakness is that a self-reported label without auditing is easy to game. An advertiser who used AI to generate a product image can simply not check the box, and Google — by its own admission — will not investigate. Consumer-protection researchers who have studied similar honor-system disclosures on social platforms have generally found compliance rates well below what regulators would consider meaningful. Google has not published how it will detect or penalize false declarations, or whether its own ad-review AI will attempt to flag synthetic imagery independently.
For the AI advertising market, the label is more consequential as a signal than as an enforcement tool. It normalizes the idea that AI-generated commercial imagery should be labeled at all, gives regulators a mechanism to point to when they push for stronger rules, and quietly incentivizes advertisers using Google's own generative tools — since only those get the automatic, credible version of the disclosure. That is a subtle nudge toward Google's ad stack, and it will likely be read as one by every rival ad platform watching the rollout.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
