Google has started rolling out AI content labels across its advertising products, and it changes a small but real part of how you ship creative. According to Google's own support documentation, the new AI label setting is launching gradually throughout July for Google Ads, Display & Video 360, Campaign Manager 360, Merchant Center and Google Ads Editor. Most write-ups filed this under AI ethics. For anyone spending money on Google Ads, it is a workflow story with a compliance edge, and it lands on the person who presses launch.
What Google actually announced
Google is adding a “How this ad was made” section to the My Ad Center panel, reachable from the three-dot menu or information icon on ads across Search, YouTube and Discover, as Search Engine Journal reported on July 13. Open that panel and it tells the viewer whether generative AI was used to create or edit the ad. The label setting that populates it is what is rolling out across the Google stack this month.
The important detail is who is responsible for the disclosure, and it splits two ways.
Google's AI labels itself. Yours does not.
If you build assets with Google's built-in generative AI tools inside Ads, the label is applied automatically. Google also embeds machine-readable provenance, SynthID watermarks and C2PA metadata, into everything its own tools generate, so those assets carry their own record. If you use anything outside Google's advertising products, a third-party image generator, an AI video editor, an AI copy tool, you have to apply the disclosure yourself through a new control. That is stated plainly in Google's documentation and confirmed in the reporting above.
Two things are still undefined, and we are not going to pretend otherwise. Google has not named which third-party tools count, and it has not defined what “created or edited” means. Whether a minor AI touch-up needs a label, or only substantially AI-made creative, is an open question Google has not answered. Until it does, the safe default is simple: if an AI tool meaningfully made or changed the asset, label it.
Where the label actually shows up
In most markets, the disclosure sits quietly inside My Ad Center, out of the main view. But for campaigns targeting the European Union, India and New York, where AI regulations require it, Google's documentation says the label also appears as a visible overlay on the ad itself. If you run in those markets, this is a mark on your creative, in front of the customer, not something buried in a menu. Google is also blunt that using its label setting does not guarantee compliance with any specific regulation, and that the legal responsibility for accurate disclosure stays with the advertiser in every case.
What this means for your account
For most advertisers the practical impact is a process change, not a media change. It does not limit your use of AI, and it does not touch bidding, budgets or targeting. The exposure is a gap in the handoff. Creative production and campaign management are often different people, or different teams, so whoever launches the campaign may have no idea a freelancer generated the hero image in Midjourney or rewrote the headlines with an AI tool. Undocumented AI use is the thing that trips this up, and it is invisible until a label is missing.
This is the same class of hygiene problem that decides whether your creative testing is even readable. If you cannot say what is in each ad and how it was made, you cannot learn from it, which is the point we made in you don't have a targeting problem, you have a creative problem. Labeling discipline and testing discipline are the same muscle.
What to do this week
Four actions, none of which need new tools.
First, inventory your live AI assets. List every image, video and headline currently in market and note which ones touched an AI tool outside Google. Ask your designers and freelancers directly rather than guessing. This is the record the whole thing depends on.
Second, turn on the AI label status column. In your Asset library and asset reporting, add the “AI label” status column, filter for assets that are still undesignated, and set each one. Google's guide walks through doing this per asset inside Asset studio, then Asset library. Thirty minutes clears the backlog.
Third, fix the handoff. Add one line to your creative brief or approval step: AI used, which tool, create or edit. Document it before launch, not after a disapproval. If an agency or freelancer supplies your creative, make them confirm third-party AI use before it goes live.
Fourth, if you run the EU, India or New York, expect on-ad overlays. Check those campaigns for the visible label and watch whether it moves click-through or conversion rate. Nobody has measured whether the overlay changes user behaviour yet, so treat it as something to monitor, not something to fear or ignore.
The honest read
This is not a fire drill. Google has not announced a firm enforcement date or whether there is a grace period before missing labels cause warnings or disapprovals, and it auto-labels output from its own tools. The accounts that get caught out will be the ones with loose creative handoffs and no record of what AI touched what. The fix is cheap: a documented AI step in your workflow, and one afternoon setting labels on the assets already running.
If you want a second set of eyes on your creative pipeline, and on whether your tracking, tagging and account hygiene are ready for changes like this, that is exactly what our free 15-point Profit Audit checks. We look at the plumbing, from signal health to how your account is actually organised, and you keep the findings either way.