Skip to main content

Google's AI labels for ads: what the July 2026 update does, and what it doesn't

Google now lets advertisers label AI-made image and video creative across its ad products. It is a useful switch. It is not a compliance guarantee, and Google says so.

AdVizion Research5 min read
A man in a cafe watching a short video ad on his phone
AI-generated photo illustration.

Key points

In July 2026 Google added ways to label AI-generated image and video ads: add a label to the creative itself, or switch on an AI label setting rolling out across Google Ads, Display & Video 360, Campaign Manager 360, Merchant Center and Google Ads Editor.

Google states that using the setting does not guarantee compliance with any specific law, and that advertisers remain responsible for local obligations such as the EU AI Act.

What changed

Google's policy update gives advertisers two routes. You can put a disclosure into the image or video yourself, or you can use the AI label setting, which Google rolled out gradually through July 2026 across its main buying tools.

Election ads are unchanged: advertisers running them must still tick the altered or synthetic content disclosure under Google's political content policy.

What the setting does not do

It does not decide for you whether an asset needs a label. That judgment, whether a viewer could take it as real, is still yours.

It does not travel with the file. The same video exported to Meta, TikTok or a programmatic exchange carries none of Google's settings. If you rely on the toggle, every other placement ships unlabelled.

And it is not a legal safe harbour. Google's own text points advertisers to legal advice for local obligations.

A practical rule for teams

Label in the creative when the asset is realistic and will run anywhere outside a single Google campaign. Use the platform setting as well, not instead. Then record which variants were labelled and why, so a later question from a client, a platform reviewer or a regulator has an answer.

Where AdVizion fits

AdVizion reviews the creative itself, not the campaign settings, so the check follows the asset into every platform you export to. Realistic AI imagery without a visible label is flagged on the ad, with the market that triggered it.

This article explains public rules for marketers. It is not legal advice. Check the source texts and your own counsel before you rely on it.

Questions

Is the Google AI label mandatory for all ads?

Google's July 2026 update describes the label as something advertisers can add, while existing disclosure rules such as election-ad disclosures still apply. Local laws may require a label regardless of the setting.

Does the label hurt performance?

There is no public Google data on that. Test it like any creative change, and do not skip a legally required label to protect click-through rate.

Sources

  1. 1.Google Ads policy: Updates to AI labelling requirements (July 2026)

Check your next ad against these rules before it spends.

AdVizion checks platform policy, claims, AI disclosure and your brand rules, and pins every finding to the ad.

See how it works

Keep reading

All posts