PALM SPRINGS, CALIF. — Excessive artificial intelligence disclosure labeling could be too much of a well-intentioned thing. Aside from stoking consumer burnout on one hand, the impact could potentially harm advertisers by highlighting AI usage that splits audience perceptions between creativity and inauthenticity.
“The risk of transparency is if you over label everything, consumers will just tune out and have a particular amount of fatigue,” Caroline Giegerich, vice president, AI & Marketing Innovation at the IAB, told Beet.TV at the IAB Annual Leadership Meeting. “There is a possibility that if you are labeling everything AI use that could actually hurt advertisers because you’re highlighting AI use.”
The IAB addressed this through nuanced guidelines focusing only on AI applications that could deceive consumers, developed by a working group including tech platforms, brands, publishers, and agencies.
Consumer deception threshold
Research revealed split consumer perceptions of AI in advertising, with some viewing it as creative while millennials and Gen Z often consider it inauthentic, prompting focus on scenarios involving potential consumer deception.
“We wanted to look at just the cases of AI use that would lead people to consumer deception,” Giegerich said, citing an example of synthetic avatars selling skincare products without disclosure that could mislead consumers expecting real testimonials.
“If I saw an advertisement for wrinkle cream, I buy the wrinkle cream, I use it for three to four weeks, I see no result, and then I realize it was a synthetic advertiser, a synthetic avatar that sold it to me — that would be consumer deception,” Giegerich said.
C2PA provides auditable layer
The IAB recommends using C2PA provenance data, which refers to the Coalition for Content Provenance and Authenticity that provides for content verification, as machine-readable “nutrition labels” that track content origins from cameras to ChatGPT, enabling business-to-business communication of AI usage.
Brands like Mondelez already implement this approach through their generative AI creative platform Ida, labeling all AI-created assets with C2PA data for digital and social use.
“C2PA is provenance data. The easiest way to think about it is the invisible nutrition label that tells you where something is from,” Giegerich said. “For them to turn on the framework that I’m suggesting is actually easier than platforms that aren’t creating things with generative AI to that degree.”
Strategy versus tool misconception
The biggest brand and agency misconception involves treating AI as an isolated tool rather than a pervasive strategy that influences measurement, commerce, video, audio, creator, and gaming initiatives horizontally.
“There’s still this perception that AI is a tool and not a strategy, and that would be a mistake,” Giegerich said. “It’s not like it’s in one place. It’s actually pervasive across all of those.”
Speed over proprietary development
Companies building proprietary AI solutions internally risk falling behind the pace of external AI advancement, making pilot-and-test approaches insufficient for achieving necessary scale.
“Unless you can move as fast as an OpenAI, let’s be honest, most of the advertising ecosystem cannot do, then you’re going to be left behind if you’re not working as fast as AI is actually moving,” Giegerich said. “The misconception is that you can just pilot, test and learn and sort of get to the scaled place that you need to be with all of this technology. I don’t think that’s possible.”
You’re watching Beet.TV coverage from IAB ALM 2026. For more videos from this series, please visit this page.
