PPCdigest.com
← Google Ads questions, answered

Can I use AI-generated images in Google Ads?

Rokas Golcas · Updated

Yes. Google Ads allows AI-generated images, whether you make them with Google's own tools or upload them from other software, but they must follow the same ad policies as any other image. If you target the EU, India or New York, local laws may require an AI label, and Merchant Center product images must keep their AI metadata.

Google's position is simple: all advertising content is subject to its ad policies regardless of how it was created. AI images go through the same reviews and enforcement as any other image, including the rules on inappropriate content and misrepresentation. So the real question is not whether an image was made with AI, but whether it is accurate and compliant.

How do you generate images inside Google Ads?

Google's image tools appear when you set up a campaign, edit an ad or asset group, work in the Asset library, and in some recommendations. You can write a prompt, or add photos of your own product so Google AI creates lifestyle images that stay faithful to it.

Prompt-based generation is limited to eligible accounts. Google says the account should:

  • Use the Google Ads interface in English, Dutch, French, German, Italian, Japanese, Portuguese or Spanish
  • Be active and have run ads for at least 30 days
  • Have a history of policy compliance
  • Not have completed Election Advertising Verification, and not advertise in a sensitive vertical such as sexual or political content

The tools are designed to block images of children, prominent likenesses, and branded items or logos you didn't provide. Every image they make carries a SynthID watermark and C2PA markup, so it can be identified as AI-generated. Generated images that you don't use in a campaign stay in your Asset library for 14 days. And Google is clear that generated assets aren't guaranteed to pass policy review.

Do you need to label AI images?

Sometimes. Google says AI regulations in the European Union, India and New York require ads with certain AI-generated or edited assets to include disclosures or labels. Since July 2026 you have two options:

  • The AI label setting. Mark an asset as "created or edited with AI" in the Asset library or when Google prompts you to review assets during campaign creation. The ad then shows a disclosure in "How this ad was made" in My Ad Center, and campaigns targeting the EU, India or New York also get a visible label on the ad.
  • Your own label in the image. Google says these labels won't break its rules against text overlays and watermarks. Keep the label away from corners and edges, and consider turning off image enhancements, which can crop it.

Google may label assets itself, for example ones created by its fully automated features, and those labels can't be overwritten. Google also says the setting doesn't guarantee legal compliance, so check the rules where you advertise. Election ads have a separate requirement: the "Altered or synthetic content" checkbox in campaign settings.

What about Shopping and Merchant Center images?

Merchant Center accepts AI-generated images in image_link, additional_image_link and lifestyle_image_link. They must contain metadata showing they were made with generative AI, such as the IPTC DigitalSourceType value TrainedAlgorithmicMedia, and you must not strip it. The main image still has to accurately show the product, with no promotional text. See whether product images can have promotional text.

My advice: use AI for backgrounds, lifestyle scenes and extra sizes and aspect ratios, not to change the product itself. If the color, shape or details in the ad don't match what the customer receives, you have a misrepresentation risk, no matter which tool made the picture. For a tour of Google's creative tools, read my look at Google's AI asset studio.

Sources