IAB publishes version two of its AI disclosure framework: labels cut click through by 31.5 percent
As legal disclosure obligations multiply, the trade body IAB updated the framework describing which uses of AI warrant a consumer facing label. The NYU finding behind it points at a clear tension: disclosure cuts an ad's click through rate by 31.5 percent.
The framework updated to the pace of regulation
Per a Marketing Dive report dated August 19, 2026, the Interactive Advertising Bureau released version 2 of its AI Transparency and Disclosure Framework.
Per the publication, the new version does not change earlier recommendations; it expands them to address a fast moving regulatory landscape. It adds guidance on using legally required labels and provides an evidence based footing for the guidelines.
A MediaPost report dated August 20 highlights the most striking data behind the framework: per research from New York University, disclosing that an ad was made with generative AI reduced its click through rate by 31.5 percent.
When a label applies and when it does not
Per MediaPost, the framework recommends disclosure when three threshold conditions are met together: a risk of deception exists, a material impact could occur, and consumer expectations require clarity.
Per the same report, the framework warns that labelling applied to AI involvement that poses no risk of deception penalizes advertisers without protecting consumers. It also notes that excessive labelling creates label fatigue and reduces long term effectiveness. In the words of IAB VP of AI Caroline Gigerentz, trust between a brand and its customers is everything and being honest about AI is part of earning it, yet not every use of AI needs a label.
The consumer picture is contradictory. Per MediaPost, more than 50 percent of consumers believe brands should disclose fully AI generated ads or AI imagery and video. Per the framework data reported by Marketing Dive, 73 percent of millennials and Gen Z say clear disclosure would increase or not affect their purchase likelihood.
The industry's own usage has risen fast. Per Marketing Dive, 83 percent of ad executives report using AI in creative processes, 23 percentage points higher than in 2024. And 72 percent of marketers believe there should be an industry standard for AI disclosure.
The reason for the update is regulation. Marketing Dive reports that the text references legal requirements in the European Union, Asia, New York and California.
Disclosure is no longer a choice, it is a design problem
The assessment in this section is ours. The number reveals a clear tension: honesty carries a measurable cost. Treating that cost as a reason to avoid disclosure would be wrong though, because in many markets disclosure is already a legal obligation.
The right reading, in our view, is this: if a label lowers clicks, the problem is not that the label exists but how it is presented and how the ad is built. The same information can be delivered like a warning or like a production note. The two do not produce the same result.
Second, the distinction about which use needs labelling genuinely matters. There is a world of difference between using AI to correct the colour of an image and presenting a person who never existed as real. The deception risk threshold in the framework draws that line.
Third, the drop is probably caused not by the label itself but by the inference the viewer makes alongside it: this ad may have been made cheaply, it may not be real. The antidote to that inference is not hiding but showing what is real. Real product, real place, real customer.
Fourth, direction. The debate about marking AI content is not limited to advertising; Anthropic adding an invisible watermark to text produced by Claude models was the technical end of the same heading. The industry is moving toward making generated content distinguishable.
What it means for businesses in Türkiye
This debate is not theoretical in Türkiye. Per Anadolu Agency, an amendment to the Regulation on Commercial Advertising and Unfair Commercial Practices was published in the Official Gazette on July 1, 2026, and part of the provisions took effect on August 1, 2026.
Per the report, where advertisements feature digital characters created with AI that cannot be distinguished from humans, this must be stated clearly, understandably and distinguishably. Advertisements in which AI generated digital versions of real people present a product experience or give a recommendation were prohibited.
The same package covers other headings too: a mandatory advertisement or promotion label for content creators, a ban on publishing user reviews from channels where purchase cannot be verified, and taking the lowest price of the preceding 10 days as the reference in discounted sales.
The assessment below is our reading. The IAB framework is not binding in Türkiye, but it answers a practical question: how do you protect effectiveness while the obligation stands? The answer runs through presenting the label as production information rather than an apology.
Three practical steps. First, record in writing where AI enters your creative production; decide on disclosure during production rather than afterwards. Second, if you use digital characters indistinguishable from humans, make that a deliberate choice from the start, because the regulation flags exactly this case. Third, measure the cost of disclosure instead of assuming it: compare the labelled version of a creative with one strengthened by real footage.
The UNALSOFT take
When producing AI advertising films our own rule is clear: the product itself is real, claims are verifiable, and the production method is not something to hide. This news also shows us the approach is not free; disclosure carries a measured click cost. Still, the move is not to dodge the cost but to build the production that lowers it: work supported by real product footage, with claims that are not inflated, telling the viewer plainly what was done. Regulation in Türkiye already points this way; the brand that adapts early comes out cheaper than the one correcting course later.
Do your ads comply with the new regulation?
Let us review the AI use in your creative process and your disclosure approach together.