Site icon Impact Digital

Google’s AI Ad Labels: A Small Advertiser’s Guide

Illustration of an ad image tagged with an AI label in an ad asset library

If you run Google Ads for a small business, or manage ads for a few clients, you have probably used AI to make or touch up an image or video. Google now has a setting for telling it which creative was made or edited with AI. This guide explains what Google’s own pages say about that setting, where to find it, how it differs from the older election-ads disclosure, and how to keep a simple record so you always know which creative was AI-assisted.

The policy details below come from Google’s own help pages, linked in the Sources section. Where Google’s text does not answer a question, this article says so rather than guessing.

What Google’s policy update actually says

Google’s Advertising Policies Help has an update titled “Updates to AI labeling requirements (July 2026)”, posted on July 9, 2026. It says the changes take effect in July 2026 and that the AI label setting rolls out gradually through that month. The key points:

Is labelling mandatory?

Third-party blogs disagree on this, and Google’s text is narrower than much of the commentary. The Google Ads Help article on the setting says laws in some regions may require labelling of certain AI-created or AI-edited ads, and advises you to check the rules that apply to you. It does not state a general, worldwide labelling requirement, and the pages cited here describe no penalties for not labelling. Beyond that, this article takes no side. If you advertise in the regions named by Google, or you are unsure, ask a lawyer rather than relying on a blog, including this one.

The AI label setting and where to find it

Google’s Help article “Use AI content label settings and disclosures” describes the setting. For each asset you choose whether to label it as created or edited with AI, or not to label it. Here is where the controls sit in Google Ads:

SituationWhat Google says to do
Creating a campaignWhen a “Review assets” prompt appears, select the assets, choose Manage AI label, pick an option, then select Done.
Existing campaignsGo to Tools > Asset studio > Asset library, open an asset’s details page, choose an option and select Save.
Generating new AI assetsIn the table, tick the checkbox for an asset whose “AI label” column is undesignated, then choose an option in the pop-up.
Checking statusUse the “AI label” status column in the Asset library and in asset reporting. You can also filter assets by “AI label”.
Uploading a video to YouTube from Google AdsYou are asked whether AI was used to generate or edit it, to meet YouTube’s disclosure policies for organic videos.

What viewers see

Ads with assets you designate as AI-generated or edited get a disclosure in the “How this ad was made” section of My Ad Center, reachable from the three-dot menu on ads across Search, YouTube and Discover. For campaigns targeting certain regions, a visible overlay also appears on the ad itself. Google’s article lists those regions as the European Union, India, New York, California, South Korea and Vietnam.

The wider rollout of visible labels to California, South Korea and Vietnam was also reported in news coverage around mid-November 2026. That date comes from industry coverage only and is not confirmed on a Google page. What is confirmed is that Google’s own Help article names those three places alongside the EU, India and New York. Reports about deepfake bans are likewise reported only: the Google pages cited here do not describe one.

Labels Google applies for you

Google may label assets itself where it is legally required to, or when it receives signals from other platforms. If you use Google’s fully automated features to create and serve creatives, assets may be labelled on your behalf. The Help article says these labels cannot be overwritten. It does not say whether you can remove a label you applied yourself, so treat your own choice as something to get right the first time.

Adding your own visible label

You can also put an AI label on the asset itself using your design tools. Google says to keep it away from the corners and edges of image assets, because rendering formats vary and trimming cannot be ruled out. It also says to opt out of image enhancements that can crop labels, naming Asset Enhancements (Display and Performance Max) and Adaptive Layouts (Demand Gen). If you use automated asset features, our guides to Performance Max sourced videos and automated promotions cover related settings worth reviewing.

How this relates to the election-ads disclosure

These are two separate things. The election-ads rule sits in Google’s political content policy. It applies to Election Ads containing images, video or audio that inauthentically depict real or realistic-looking people or events, such as making someone appear to say or do something they did not. Trivial edits like resizing, cropping, colour correction or red-eye removal are out of scope.

Election advertisers tick a checkbox in the “Altered or synthetic content” section of campaign settings. For feeds and Shorts on mobile phones, and in-stream ads on computers, phones, mobile web and TV screens, Google generates the in-ad disclosure from that checkbox. For other formats the advertiser must also add a clear and conspicuous disclosure. Google’s July 2026 update tells election advertisers to keep using that checkbox, so the new AI label setting does not replace it.

Election ads disclosureAI label setting
Who it concernsElection advertisersAdvertisers using AI-generated or edited image and video assets
Where you set it“Altered or synthetic content” in campaign settingsAsset library or “Review assets” step
LevelCampaignIndividual asset

A workflow for tracking AI-made creative

Google’s setting works asset by asset, so the practical risk for a small team is forgetting what was touched by AI. A shared spreadsheet is enough. This is a suggested process, not something Google requires.

  1. Log at creation. When a designer, freelancer or tool makes an asset, record it the same day: file name, date, who made it, and whether AI generated or edited it.
  2. Record the tool and what it did. Note whether AI created the whole image, extended a background, swapped an object or only cleaned noise. Google’s pages do not define “AI-generated or edited”, so a written note helps you make a consistent call or ask a lawyer.
  3. Ask suppliers. Add a line to freelancer briefs requiring them to say whether AI was used.
  4. Set the label before launch. At the “Review assets” step, use Manage AI label, and record the choice in your sheet.
  5. Audit monthly. Filter the Asset library by “AI label” and compare it with your log. Look for undesignated assets and any labels Google added automatically.
  6. Keep targeting in view. Note which campaigns target the regions where Google shows visible overlays, and check how your own labels survive cropping. Our guide to Google AI Max is useful if automation is generating assets for you.

Common mistakes

Key takeaways

FAQ

Do I have to label every AI-assisted image?

Google’s Help article does not state a general requirement. It says laws in some regions may require labelling and tells advertisers to check the rules that apply to them. Seek legal advice for your situation.

Will the label appear on my ad?

A disclosure appears in “How this ad was made” in My Ad Center. For campaigns targeting the EU, India, New York, California, South Korea and Vietnam, Google says a visible overlay also appears.

Does the label replace the election ads checkbox?

No. Google tells election advertisers to keep using the “Altered or synthetic content” checkbox in campaign settings.

Can I add my own AI label to the image?

Yes. Google says such labels will not violate its text overlay and watermark policies. Keep them away from corners and edges and consider turning off image enhancements that can crop them.

Sources

Exit mobile version