Digital Marketing

Google Ads AI Label: The "How This Ad Was Made" Panel Explained

Talha Aslan 18 min read 4 views

Short answer

What is the Google Ads AI label?

The Google Ads AI label is a disclosure that tells people an ad asset was created or edited with AI. Users read it in the three-dot menu on an ad, in a section called "How this ad was made." In some regions, a visible label also appears on the ad. Google adds it automatically, or you set it.

Keep reading for the full details and examples

So the Google Ads AI label is part of a wider push for ad transparency. In short, people who see an ad can learn how it was made. As an advertiser, you need to know which of your assets carry a label and manage them when needed.

Why does this matter? AI-generated images are now an ordinary creative tool. Besides, users are curious about how ads get made. Clear information is the simplest way to protect trust, and the Google Ads AI label is the advertising side of that.

In this guide, we cover the panel, the label setting, regional rules and the team process. We also show where our free responsive search ad builder fits into your workflow.

We wrote this as a lasting reference. Screen names and region lists can change, so verify current details on Google's official help pages.

What does the ad transparency panel show users?

According to the Google My Ad Center help page, "How this ad was made" is one of the options in the three-dot menu on an ad. There, a user then sees how the ad was created and whether AI helped create or edit it. The label can read differently depending on your interface language.

Also, the same menu includes "About this advertiser," "Why this ad" and "Report this ad." As a result, a user can inspect the ad, the advertiser and the targeting in one place.

One detail matters here: the section appears only on ads that use AI-edited or AI-created assets. So most ads will not show it. In other words, the presence of the panel is a hint that the ad contains an AI asset. However, a missing panel proves nothing, because labeling depends on the advertiser and on Google's automatic rules.

People can also find post-click label information in My Ad Center and in machine-readable metadata (SynthID and C2PA). For details, see the My Ad Center help page.

Where does the Google Ads AI label appear?

According to Google's official blog post, the transparency panel is accessible for ads on Search, YouTube and Discover. In other words, a user who sees an ad on those surfaces can open the panel. For the full statement, read Google's official blog post.

In other words, the label appears only on ads that contain labeled assets. If none of an ad's assets were created or edited with AI, the panel shows no AI information.

Here is something you should know as an advertiser: you are not the only one who can add the label. Google may label assets itself when the law requires it or when it receives signals from other platforms. With fully automated features, Google may label assets on your behalf.

So do not limit label management to the assets you mark yourself. Check the label status in your asset library on a regular basis. For example, a teammate may have turned on an automated feature, and then the system generates and labels new assets without anyone noticing. A monthly look prevents such surprises, and filtering by label in reports speeds up the work.

Who adds the label: Google or you?

There are three routes. The first is Google's own generative AI ad tools. For assets you create with those tools, Google adds the disclosure to the My Ad Center panel automatically. You do not need to do anything extra.

The second route covers assets you prepare with other tools. If you create an image or video with an outside tool, you mark it yourself with the AI label setting in Google Ads. Google cannot know on its own whether that asset came from AI. Therefore the accuracy of the declaration is your responsibility. This matters most for files from outside designers, because the creator is usually the only one who knows how a file came about.

The third route is Google's own decision. Google may apply a label when the law requires it or when an outside signal calls for it. You cannot overwrite a label that Google applies.

All three routes can run in the same account. For example, one campaign may hold assets from Google's tools and assets you imported. Keeping a source record for each asset saves time when questions come up, and it keeps knowledge in place when your team changes.

Which assets can you mark with the AI label setting?

According to Google Ads Help, the label setting covers image and video assets. The help page describes AI tools that create new assets from prompts or enhance existing ones. You find the setting not only in Google Ads but also in Search Ads 360, Display & Video 360, Campaign Manager 360, Merchant Center and Google Ads Editor.

However, text assets such as headlines and descriptions do not appear on that page. So check the current scope on the official page for text. Even so, keeping an internal record of AI-written text is a good habit.

For image and video assets, consider labeling these:

  • Images or videos you generate from scratch with AI.
  • Existing images that AI extended, completed or heavily changed.
  • AI assets that realistically depict a person, place, object or event.

The page says an advertiser can disclose an asset as AI-generated or AI-edited, including when it realistically depicts a person, place, thing or event. In those cases, labeling is a simple way to treat users honestly.

For small edits such as a crop or a color fix, read the help page before you decide. Not every edit weighs the same. For instance, redrawing an image entirely with AI is not in the same category as resizing it, so set a threshold in your team and write it down.

How do you manage the Google Ads AI label setting?

According to Google Ads Help, you use the setting in three places. Screen names can change over time, so treat the steps below as a guide and verify the current path on the help page. For details, see the AI label help page.

  1. For existing campaigns, go to the asset library, open the asset's details page, choose the label option and save.
  2. During campaign creation, if the system asks you to review assets, select the assets, open the label management option and make your choice.
  3. When you generate new AI assets, use the label column or checkbox in the table.

You have two options: label the asset as created or edited with AI, or choose not to label it. If you upload a video to YouTube from within Google Ads, the system also asks whether you used AI to create or edit it.

In the asset library and in asset reports, you can see label status and filter by label. That view is the easiest way to see which assets carry a label across your account.

If you do not want to track these settings one by one, our google ads management work includes asset labels in the process. First, though, we suggest building your own asset inventory. An inventory is a simple table that shows which image sits in which campaign and where it came from. Without it, using the label setting correctly gets harder.

In which regions does a visible label appear on the ad?

According to Google Ads Help, for campaigns that target the European Union, India, New York, California, South Korea or Vietnam, ads with AI-created or AI-edited assets also show a visible overlay. Because the list may change over time, check the current list on the help page.

The visible overlay differs from the My Ad Center disclosure. For example, the panel appears when a user opens the menu, but the overlay sits directly on the ad. So design your visuals knowing the overlay will take up space, and keep key product details or call-to-action text away from the edges, because overlays and cropping can affect edge areas.

Here is a sample scenario. Say you run a campaign that targets both the European Union and California. Both appear in the help page's list, so ads with AI assets may carry an overlay. In that case, designing the creative away from the edges is the safest path, and one design then fits both regions. This is an example scenario, so check the current rule. As the region list grows, designing every creative with safe margins from the start saves you from rework.

None of this is legal advice. Google itself states that using the AI label setting does not guarantee compliance with specific regulations. For regulation questions, get legal help.

What should you watch when you add your own label?

Google Ads Help allows you to add a visible AI label with your own design tools. According to the page, such labels do not violate Google's policies on text overlays and watermarks. Two practical tips apply.

First, do not place the label close to corners and edges. Edges can get trimmed when the system renders the image, so your label may get cut off.

Second, review the image enhancement features. The help page says they can crop your label. The relevant controls are asset enhancements in Display and Performance Max campaigns, and adaptive layouts in Demand Gen campaigns. If you want to protect your own label, consider turning those features off.

Your own label does not replace Google's label setting. They are separate things. Your label is part of the visual, whereas the setting manages the My Ad Center panel and regional overlays. Therefore consider them together and keep them consistent.

For example, if the corner of an image says "made with AI," mark the same asset in the setting as well. Inconsistency raises questions for users and for reviewers alike.

What are SynthID and C2PA, and what do they do here?

SynthID is a non-visible digital watermark that Google embeds in AI-generated content. C2PA, short for Coalition for Content Provenance and Authenticity, is a standard for markup that records where content came from. Both work through machines, not through human eyes.

According to Google Ads Help, Google embeds both signals in all images and videos you generate with Google Ads tools. They are not labels that users see. They are machine-readable provenance information.

So think in two layers. First, there is what users see: the panel and, in some regions, the overlay. Second, there is the machine-readable metadata in the background. In short, the first layer speaks to people, and the second speaks to systems. Your main responsibility sits in the first layer, where you declare things accurately.

We covered C2PA in more depth in what C2PA content credentials are. For copyright questions about AI images, read can you use AI-generated images commercially.

What changes for election ads?

Election-related ads carry a separate rule. According to the My Ad Center help page, for election ads with consequential synthetic or altered content, the advertiser must select the "Altered or synthetic content" checkbox in the campaign settings.

Some formats also require a prominent disclosure inside the ad. So the panel disclosure alone may not be enough.

If you do not run election ads, this section does not concern you directly. Still, the existence of the rule shows how seriously Google treats synthetic content. Take extra care whenever AI-generated content imitates real people or events. A fake image that resembles a real person or event can mislead users even with a label, so avoid such visuals.

Google's official blog post adds one more reminder: ad policies that prohibit misleading ads apply whether or not AI is involved. In other words, a label does not make misleading content acceptable. For the wider picture on AI content, see our piece on whether Google penalizes AI content.

Does the label guarantee compliance?

No. Google Ads Help states plainly that using the AI label setting does not guarantee compliance with specific regulations. In some regions, laws may require labels on certain assets that AI created or edited. Ask a lawyer which rules apply to you.

So this article is not legal advice. We only summarize the process that the official help pages describe, from an ad manager's point of view. In other words, the summary offers a practical workflow rather than a legal reading.

Also, the label does not cancel the rest of the ad policies. A misleading claim, an unrealistic promise or a fake image is a policy problem even when it carries a label. So do not treat the label as a "shield." It adds honest information next to correct content, and it does not fix wrong content.

The healthiest approach has three steps. First, make sure the content reflects reality. Next, add the right label. Finally, check regional rules with legal support.

How does label status change by scenario?

The table below summarizes who adds the label, and how, in common scenarios. It is a general frame that we compiled from the official help pages. The exact requirement depends on your region and your content.

ScenarioLabel statusWho marks itNote
Asset made with Google's generative AI ad toolDisclosure is added automaticallyGoogleNo extra step needed
Image or video made with an outside toolYou mark itAdvertiserUse the label setting
AI asset that realistically depicts a person, place or eventMarking is a good ideaAdvertiserThe help page mentions this case
Legal requirement or outside signalGoogle may label itGoogleYou cannot overwrite the label
Asset from a fully automated featureMay be labeled for youGoogleCheck status in asset reports
Election ad with synthetic contentSelect the campaign checkboxAdvertiserSome formats need a prominent disclosure
Text asset (headline, description)Not covered on the help pageAdvertiserVerify the scope from a current source

Does the label affect ad performance?

We found no official data on this question. So we cannot say that "the label lowers clicks" or that "it has no effect." So instead of writing a number, we suggest you measure it in your own account.

Set up an experiment to measure it. Show different creative to the same audience with the same message. Still, do not switch off a label that may be required for the sake of a test. Run the experiment only with assets where the labeling decision is yours, and within the official rules.

For experiment setup, see our guide to Google Ads drafts and experiments. To read the outcome, our free ab test calculator shows significance, confidence intervals and a sample ratio check.

Remember that skipping the label for performance reasons creates rule and trust risks. A small short-term difference does not justify long-term damage to your reputation. Besides, a brand that uses labels correctly looks open and honest to users.

How do you set up label management in your team?

Instead of remembering labels one by one, build a simple process. The process has three steps.

  1. Record the source of every asset: Google tool, outside tool or human-made. A simple table is enough.
  2. Give one person approval responsibility before launch. Check the label setting at the same point as design approval, and write the source and label decision into the approval note.
  3. Review the label column in the asset library once a month. If an AI asset has no label, fix it. That way you also catch assets added later.

If you work with an outside agency or a freelance designer, ask about AI use in the contract or brief. Repeat the question for each new job, because tools and habits change. You publish the asset, so the declaration is your responsibility. If the designer does not tell you about AI use, you may publish an asset that should carry a label.

Also build a shared vocabulary. Everyone should know the difference between "I made it from scratch" and "I edited it with AI." That way, the labeling decision does not change from person to person. New team members pick up the process quickly, and a short internal document keeps that vocabulary alive. In a small business, the process can be one shared sheet. In a large team, it can connect to an approval tool.

How do AI assets fit different campaign types?

AI assets show up in different ways across campaign types. Knowing where assets come from in each type makes the labeling decision easier.

In Performance Max campaigns, the system combines and can enhance assets. For setup details, see how to set up Performance Max for ecommerce. Demand Gen campaigns lean heavily on visuals, and our Demand Gen campaign guide explains that type.

Also, in Search campaigns, AI-assisted features can expand. Our piece on Google Ads AI Max for search campaigns covers that topic. Automated recommendations can also suggest assets and settings, and you can read how to control them in how to turn off auto-apply recommendations.

One rule stays the same in every type: review the assets that automated features create in your asset reports. With fully automated features, Google may label assets on your behalf. Therefore knowing what the system produces for you matters for both brand safety and label accuracy.

How can the responsive search ad builder help when you draft ad text?

If you wrote headlines or descriptions with AI help, collect and check the text before it goes live. Open our google ads preview tool, then enter the final URL and the display paths. Next, add your headlines and descriptions.

The tool shows a 30 character limit for headlines and a 90 character limit for descriptions. It also flags policy issues such as exclamation marks, all-caps words and emojis. Desktop and mobile preview tabs show how the ad may look. Finally, you can export the draft for Google Ads Editor.

This tool does not manage the label setting itself, because the setting covers image and video assets. Still, marking AI-written headlines in a list strengthens your team record. Sharing the drafts you build in the tool also speeds up approval. To review pinning, see pinning headlines in responsive search ads.

Make sure the text reflects reality. AI can invent a claim, and a made-up feature or offer is a policy problem whether or not the ad carries a label. So verify every claim before launch.

Which mistakes show up most often in labeling decisions?

These mistakes repeat in many accounts. Check your own process against the list.

  • Uploading an AI-made outside asset without a label. Google cannot know the source for you.
  • Leaving every asset unlabeled and assuming the panel will not appear anyway.
  • Trying to change a label that Google applied. You cannot overwrite it.
  • Placing your own label near the edge, where cropping cuts it off.
  • Treating the label as a compliance guarantee. Google rejects that reading.
  • Skipping the rules for your target regions.
  • Producing misleading content and trying to excuse it with a label. Policy prohibits this.

Most of these mistakes come from a missing process, not from a lack of knowledge. So a source record and an approval step are the most effective safeguards. A regular review calendar also helps you spot the mistakes early.

Which checklist should you run before launch?

Run this list in order for every new campaign. Each item matches a section above.

  1. Do you know the source of every image and video in the campaign (Google tool, outside tool, human-made)?
  2. Did you mark the AI-created or AI-edited assets with the label setting?
  3. Did you check the labels that Google applied automatically in the asset library?
  4. Do your target regions have an overlay rule?
  5. Do key text and logos sit away from the edges?
  6. If you run election ads or realistic synthetic content, did you select the related setting?
  7. Does the content reflect reality, and is any claim misleading?
  8. Did you ask a legal expert when the rules required it?
  9. Did you write the source record and the approval into the team sheet?

If any item says "no," fix it before you publish. Because each item is a short check, adding the list to your workflow takes little time.

What if you cannot set up the label and account process alone?

For many businesses, this process is manageable. Recording asset sources, marking the label setting and checking once a month is usually in your hands.

If your account has many campaigns, several countries or assets from outside agencies, the process can get tangled. Besides, each country's rule may differ, and tracking them takes time. In that case, Talha Aslan and our team can help you build your asset inventory and label process together with your account structure.

To see how we set up, monitor and keep improving ad accounts, review our google ads agency services page. For consistency between your ad and your landing page, our sibling article on ad to landing page message match also helps.

No team can guarantee a specific result. We only help you record your assets correctly, manage labels on purpose and base your decisions on official sources. As the rules can change quickly, review the process at regular intervals. This article is not legal advice.

Core guide: What Is Digital Marketing? Its types, examples and how it is done, on one page with an interactive film that travels through the brain.

Frequently Asked Questions

Is the AI label mandatory in Google Ads?
Not for every piece of AI content. According to Google's help pages, requirements vary by region and content type, and they can differ across Google platforms. Some regions do have labeling rules. So check the rules for each region you target. Nothing in this article is legal advice, and you should consult a lawyer for compliance questions.
Will I see the "How this ad was made" section on every ad?
No. According to the Google My Ad Center help page, the section appears only on ads that use AI-edited or AI-created assets, so most ads do not show it. A person reaches it through the three-dot menu or the info icon on an ad. Therefore, a missing panel does not prove an account has no AI assets.
Do I need to label assets I made with Google's own AI tools?
Usually not. Google adds the disclosure to the My Ad Center panel automatically for assets you generate with its own generative AI ad tools. Still, check the label status in your asset library from time to time. Also remember that you cannot overwrite a label that Google applies itself.
How do I mark an image I made with another tool?
Use the AI label setting in Google Ads. In the asset details, or in the step where you review assets during campaign setup, you will find an option to label the asset as created or edited with AI. Screen names can change, so verify the current path on the Google Ads Help page.
Will the AI label hurt my ad performance?
Google has published no official data on this, so we cannot say for sure. Measure the effect in your own account. Compare labeled and unlabeled creative on the same audience in a proper experiment, and then read the result with a significance check. Skipping the label to protect performance can create rule and trust risks.
How do SynthID and C2PA affect my ads?
According to Google Ads Help, Google embeds a non-visible SynthID watermark and C2PA markup into images and videos you generate with Google Ads tools. These are not labels that people see. They are machine-readable provenance signals. You do not need to take any extra action, but share this fact with your team.
  • google ads
  • ai label
  • ai transparency
  • my ad center
  • synthid
  • c2pa
  • ad policy
Share:
Talha Aslan

Google Partner digital marketing expert. Hands-on with SEO, Google Ads, web design and e-commerce projects since 2012; every post here comes from that experience.

Next project

Get a quote for your project.

Your brief goes straight to Talha Aslan and team: strategy led by Talha, delivery by an experienced team. We reply the same day and send the scope, timeline and price as a written quote within 1 business day.