Not everything AI touches needs AI ad labeling. At least, that’s the position the IAB is taking.
The ad industry trade group has released Version 2 of its AI Transparency and Disclosure Framework, with updated guidance for when advertisers, agencies, publishers and platforms should disclose the use of AI in their campaigns.
This doesn’t deviate much from the initial framework unveiled back in January. But we do get more concrete, granular guidance, especially now that some AI disclosure regulations have gone into effect.
Why This Matters:
The IAB is continuing to push for more of a targeted disclosure system rather than labeling everything created or touched by AI.
Under their framework, disclosure is recommended when AI could “materially” affect authenticity, identity or representation — meaning if consumers could be misled about whether what they’re seeing is “real.” That includes fully AI-generated images and video, some synthetic voices, synthetic avatars, digital twins, and AI assistants that consumers could mistake for actual people.
But plenty of AI uses wouldn’t automatically trigger a disclosure, including routine post-production, internal workflows, background music, generic synthetic voices, text and copy, clearly fantastical imagery, and obvious cartoon or stylized avatars.
In the US, advertisers following the framework can either use a standardized sparkle icon or add clear text to disclose AI involvement. (Keep in mind, there are also backend ways to identify AI-generated content. This is the consumer-facing piece, however.)
Of course, the update comes as AI disclosure has become a regulatory hot button, with new requirements rolling out in California and New York, and even more regulation overseas. (See the controversy over Claude watermarks.)
Experts React:
IAB says over-labeling could ultimately be counterproductive.
“Not every use of AI needs a label — labeling everything teaches consumers to ignore labels and could negatively impact advertisers,” says Caroline Giegerich, VP of AI at IAB.
At the same time, some studies have found that AI ad labeling can actually improve the relationship with consumers, leading to better brand perception and even performance lift. Transparency = authenticity, essentially, when it comes to AI ads, with all boats rising.
Our Take:
This feels like a good call on the IAB’s part. As AI gets baked into practically every part of advertising, slapping an “AI” label on anything AI touches is both onerous and meaningless.
The harder question — and the one the IAB is trying to answer here, we think — is when AI meaningfully changes what a consumer believes they’re seeing or interacting with. Focusing disclosure on the more obvious, commonsense cases makes the most sense.
Will everyone do it, though? There’s a crop of performance marketers that doesn’t seem particularly concerned with disclosure. So, for any of this to stick, enforcement will ultimately need to have some teeth.
