What an 'AI-Generated' Ad Label Actually Tells You: Almost Nothing

A 'Made with AI' tag can cover everything from a minor retouch to full machine generation. Cardiff and King's College researchers say the label obscures as much as it reveals.

Wire notes

  • Research from Cardiff University and King's College London finds AI disclosure labels indicate how an ad was produced but not the balance of human and machine contribution.
  • Evidence shows disclosing AI involvement can reduce consumer trust in advertising.
  • Meta replaced its 'Made with AI' label with 'AI info' after admitting it appeared on content with only minor AI modifications.
  • The EU is imposing strict disclosure requirements on providers of certain generative AI systems.
  • Coca-Cola and Gucci faced criticism over AI-led campaigns.
What does an ‘AI‑generated’ label in advertising actually tell us?
PhotoWhat does an ‘AI‑generated’ label in advertising actually tell us? — Pexels / pixabay

By Amara Osei

A label reading "AI-generated" tells consumers an ad was made with machine help — and nothing about how much, according to research from Denitsa Dineva, Reader (associate professor) in Marketing at Cardiff University, and Oguz A. Acar, Professor of Marketing and Innovation at King's College London.

The label covers two extremes at once. AI may have served as one editing tool inside a largely human-led creative process — refining text or altering an image. Or it may have produced most of the finished advert with limited human input. "A simple 'AI-generated' label can encompass both extremes and everything in between," the researchers write.

Why the ambiguity matters for marketers

The distinction is not academic. The researchers point to evidence that disclosing AI involvement can reduce consumer trust in advertising. Other research documents broader consumer resistance to AI, though reactions vary depending on what the technology creates.

The reputational risk is already visible. AI-led campaigns from Coca-Cola — including its AI-generated Christmas ad that drew renewed backlash in November 2025 — and Gucci attracted public criticism. A label that invites judgment about "who or what created the content" can sharpen that scrutiny rather than defuse it.

The takeaway: vague disclosure converts a technical detail into a trust liability.

How is AI disclosure different from '#ad'?

Traditional advertising disclosure is well established. Influencer hashtags such as "#ad" and "sponsored content" labels tell consumers why content appears in front of them — its commercial intent.

AI labels do something different, the researchers' own published work finds. They signal how content was produced. And there the useful information stops. For decades, designers retouched photographs and removed backgrounds while keeping creative control. AI blurs the line between editing existing content and creating new content, because the system can generate elements on its own while the human-to-machine contribution ratio shifts case by case.

The label, in short, answers a different question than the one consumers actually ask.

What are platforms and regulators doing?

Two parallel fixes are emerging, each with caveats.

  • Regulation. The EU wants providers of certain generative AI systems subject to strict disclosure requirements.
  • Provenance technology. Standards such as C2PA Content Credentials let digital content, including adverts, carry accessible information about its origin and editing history.

Meta's experience shows how hard the wording is. The company acknowledged its "Made with AI" label appeared on content involving only minor AI modifications and lacked sufficient context. It renamed the tag "AI info" and later made the label less prominent when AI was used to edit rather than generate content.

The mirror image is now appearing too. "Human-made" and "no AI" labels are emerging to signal human authorship — and they carry the same ambiguity. Does "no AI" mean no AI whatsoever, or a human process where AI played a supporting role? "Both kinds of label can obscure as much as they reveal," the researchers write.

The lesson: granularity, not vocabulary swaps, is the unsolved problem.

What should brands and consumers take from this?

For consumers, the researchers offer a measured position: "AI-generated" does not mean the result is fake, poor quality or untrustworthy. Nor does it specify the human role — who developed the idea, selected and refined the output, checked accuracy and decided what appeared.

Their closing argument applies equally to brand teams weighing disclosure strategy: an AI label "should be a starting point for evaluating content, rather than the end." Until labels communicate the degree of human judgement behind a campaign, brands relying on the binary tag are disclosing less than they think.

via digitalinformationworld.com (Source)

Filed under

Share this article:

More from Amara Osei

Amara Osei

Show full bio

News editor covering people, launches and day-to-day practice in marketing, advertising, marketing technology, media, SEO and search at Marketing Herald.

346 articles

More on the wire

« Previous articleNext article »