AI won its place in ad creative on economics alone. 83% of ad executives say their company now uses AI somewhere in the creative process, up from 60% in 2024. Nobody needed convincing.

Trust is the part economics never bought. 73% of consumers say knowing an ad was made with AI would either increase their likelihood to purchase or make no difference at all, yet fewer than half of advertisers disclose it. The industry is paying a trust cost to avoid a disclosure most people would shrug at.

That is the thinking behind the AI label we rolled out in August.

What is the Adskeeper AI label?

The AI label is a disclosure marker attached to any creative generated fully or partially with the AI tools available in the Adskeeper system. It appears on the ad as an AI ⓘ icon. Clicking it opens our DSA landing page for EU users, which explains that the content was generated fully or partially using AI.

Native ad creative showing the Adskeeper AI ⓘ label in the top corner of the ad block
The AI ⓘ label as it appears on a native ad block.

It states a production fact and nothing else. It also puts us on the right side of Article 50 of the EU AI Act, which requires AI-generated or manipulated content to be clearly marked and detectable in machine-readable form.

How does the AI label build trust?

By replacing a suspicion with a fact. Readers already assume AI is in the mix: 71% of Gen Z and Millennial consumers believe they have seen an AI-generated ad. What they have never had is a straight answer about which one, and assumption is always harsher than disclosure.

The opportunity sits in the distance between what the industry believes about its audience and what the audience actually thinks. IAB and Sonata Insights surveyed 505 consumers and 104 industry executives between October 2025 and January 2026 and found a 37-point spread between the two. Closing that spread has been treated as a branding problem for two years. It is a labeling problem, and labeling is cheap.

There is a second effect that compounds. When every AI creative carries the marker, the marker stops signaling anything unusual. It becomes ordinary, like a sponsored tag, and it does a quietly useful job: answering a question the reader was going to ask anyway, before they answer it themselves and get it wrong.

For a publisher, that lands where it matters most. Inventory that explains itself reads as a site that respects the people reading it.

What creatives get an AI label?

Everything produced with an in-platform AI tool, whether an advertiser made it or our own design team did.

Creative typeHow it gets labeled
ImagesGenerated or edited with in-platform AI image tools: CTR Guard automatic and manual generations, image generators and the editor inside the creative creation form
Motion creativesLabeled when generated from a static creative
TextGenerated with the in-platform AI text tool, for AI-impacted categories only
Internal design team assetsLabeled automatically based on the saved image source when our designers use AI tools
Rich MediaLabeled via a checkbox set by the designer building the asset

Setup work is zero. Advertisers get no extra field, no checkbox and no second creative version to maintain, and bulk uploads behave the same way. Publishers install nothing: the label travels with the creative through moderation and delivery, with no new tags, widget updates or code changes.

Which text categories fall under AI labeling?

Text labeling applies to AI-impacted categories, the ones where accuracy carries real consequences for the reader. Three groups:

  • Health and medical: Alternative Medicine, Blood Disorders, Cancer, Diabetes, Infectious Diseases, Medical Services, Sexual Health, Skin Care, Weight Loss
  • Financial: Financial Assistance, Insurance, Options, Personal Debt
  • News and social: General Social, News and Politics

What does the AI label mean for publishers?

It means the trust layer on your inventory gets handled at network level, without a revenue trade-off.

What you getWhat it does for your site
Automatic AI disclosure across images, motion, text, Rich MediaMeets Article 50 marking expectations without dev work or tag changes
In-house creatives labeled on the same termsNo exemption for network-produced assets running on your pages
DSA landing page for EU usersThe explanation lives with us, not on your site
Same moderation standard on every creativeLabeling adds information, it does not replace review
Full demand pool, unchangedNo advertiser is penalized in the auction for using AI tools

Your fill and your payouts are not the price of doing this properly. The label adds an information layer to creatives that were already going to run, and it does that on the side of the trade where readers form their opinion of your site.

To see what is currently serving on your widgets, check your publisher dashboard. For format-level detail on how these creatives appear in placement, see the native ad formats overview.

Frequently asked questions

When did AI labeling start on Adskeeper?
1 August 2026, in line with the EU AI Act’s Article 50 transparency obligations, which came into application on 2 August 2026.
Does the AI label appear on every ad?
No. It appears on creatives generated fully or partially with in-platform AI tools. Creatives made without AI carry no label.
Is AI-generated text always labeled?
No. Text labeling applies to AI-impacted categories across health, financial and news and social verticals. The full list is in the section above.
Do publishers need to change anything to support the label?
No. The label is handled at network level and travels with the creative. No new tags, widget updates or code changes.
Does labeling reduce advertiser demand on my inventory?
No. Advertisers are not penalized in the auction for using AI tools, and labeled creatives compete on the same terms as any other creative.
What does a user see when they click the AI ⓘ icon?
EU users are taken to our DSA landing page, which explains that the content was generated fully or partially using AI.
Are creatives made by Adskeeper’s own design team labeled?
Yes. Visual assets produced internally with AI tools are labeled automatically based on the saved image source, and Rich Media is labeled by the designer building the asset.