Future of Marketing Briefing: The ad industry has an AI label problem

Which is to say the question of where to draw the line is inseparable from the question of how much marketers are prepared to lose by answering honestly.  

“How do marketers react to that, knowing that there is a lot of angst in the marketplace right now around AI and its impact on society and the economy,” said Nada Bradbury, CEO of AD-ID.

The industry is trying to work that out. Trade bodies are developing guidance, brands are drawing their own lines and regulators are being lobbied for something more proportionate than the blunt instruments currently on the table. But the process is slow and the clock isn’t. The European Union’s AI transparency provisions hit their compliance deadline in August. New York state requires disclosure of AI-generated humans in marketing from June. The lines the industry hasn’t finished drawing are about to be drawn by people with different priorities entirely. 

“Over the past couple of months alone, I must have had a good five, six or seven brands reach out asking, ‘do we have guidance?’ and ‘do we have insights into how brands are thinking about this issue?’,” said Gabrielle Robitaille, director of policy and AI community lead at The World Federation of Advertisers.   

The reason, she said, is the same one that drives most compliance conversations: proximity. With August approaching, legal teams that had previously left the question to marketers are now demanding answers, and marketers are looking for somewhere to turn. Anyone who lived through the arrival of the General Data Protection Regulation will recognize the feeling. The question now is what, exactly, marketers are supposed to do with the time they have left. 

The WFA’s answer is deliberately narrow. The starting point, Robitaille said, is that the fundamental principles of advertising self-regulation still apply regardless of what technology is being used. Don’t use it to mislead. Don’t use it to generate unsubstantiated claims or inflate product efficacy. If you stripe out all those potentially harmful use cases, she argued, the remaining grey area is actually quite small. 

What’s left is a matter of threshold. The WFA’s position is that labeling should kick in when AI is central to the commercial nature of the message – i.e. when it materially shapes what someone believes about a product. 

Below that line, disclosure is a judgement call. Above it, the responsible move is to label. The clearest case, Robitaille said, is synthetic humans, where public unease is consistent and well-documented across research. A generated beach backdrop behind a shampoo bottle is a different matter entirely – technically AI but unlikely to affect anyone’s decision to buy. 

The WFA isn’t alone in trying to work this out. In January, the IAB published what it billed as the industry’s first unified standard for AI disclosure in advertising – and arrived, broadly, at the same destination. Don’t label everything. Focus on content that could genuinely deceive. Treat synthetic humans as the clearest trigger. Background alterations, audio enhancements and post-production tweaks – all fall below the line under what the IAB calls “standard production techniques”. 

Where it goes further is in the specifics. Digital twins of living people depicted in fabricated events must be labelled as well as prompt-driven AI images and videos. Deceased individuals rendered by AI require disclosure,  IAB also requires C2PA metadata to travel with every ad — so platforms like Meta, TikTok and YouTube can see exactly what AI was used and apply their own label if needed. 

“Advertisers’ response to the framework has been thoughtful and pragmatic,” said Caroline Giegerich, vp of AI at the IAB. Most brands aren’t looking to hide their use of AI and instead are looking for clarity. The real question brands are grappling with isn’t ‘Should we disclose?’ but ‘When does disclosure meaningfully serve the consumer versus when does it create confusion or fatigue?’ Right now, there’s understandable sensitivity around labeling something as AI-generated because the term can carry unintended connotations from inauthenticity to manipulation.”

If the term AI carries those connotations, some brands have drawn the obvious conclusion: don’t associate yourself with it at all. Enter the “No AI” disclaimer – advertising’s certified organic label. 

Take Aerie, the American Eagle-owned intimates brand. It has cast Pamela Anderson in an ad to promote a commitment it had already made last October not to use AI generated bodies or people. In it, the actor is shown prompting a chatbot to create models before revealing they were real people all along. That promise is itself an extension of the brand’s 2014 pledge not to retouch people in its ads. For Aerie, “No AI: isn’t a compliance decision but its brand DNA. 

Baby products brand Coterie has gone further still, committing publicly to using no AI-generated images in its social media marketing at all – a stance its CEO Jess Jacobs told the Wall Street Journal was a trust play in a crowded market where parents are a particularly skeptical audience. Le Creuset has been going out of its way to clarify that its recent social content – visually inventive videos by digital artist Ian Padgham, in which the brand’s signature cookware is transformed into unlikely objects – involves no AI whatsoever. The clarification comes in the comments of those posts, unpromoted, before anyone has a chance to assume otherwise. 

It’s a remarkable turn for an industry that spent three years telling clients AI would transform everything. 

“Over time, consistent transparency will normalize AI’s role in advertising, just as we’ve normalized other technologies before it,” said Giegerich. “The industry’s goal shouldn’t be to over-label or under-label — it should be to disclose in a way that informs without overwhelming and builds confidence rather than doubt.”

Numbers to know

$30 billion: Anthropic’s reported run-rate revenue

$102 billion: Total ad revenue OpenAI is forecasting by 2030

61%: Percentages of U.S. retail business decision makers who use media mix modelling to measure incrementality

78%: Percentage of U.S. millennials (not Gen Z) that are more likely to use a second screen during the 2026 World Cup matches

What we’re reading

Why Anthropic’s new model has cybersecurity experts rattled

Anthropic has effectively said that the model is too dangerous to release to the general public, because models of this caliber could soon be accessible to criminals, hackers and nation states, per Platformer.

OpenAI asks California, Delaware to investigate Musk’s ‘anti-competitive behavior’ ahead of April trial

OpenAI has sent a letter to California and Delaware attorneys general, insisting they investigate the anti-competitive behavior by Elon Musk and his associates, according to CNBC.

Why a podcast acquisition is actually the first step toward building an AI advertising stack

The AI Ad Economy breaks down how OpenAI’s acquisition of TBPN supports the AI platform’s advertising ambitions.

Perplexity monthly revenue jumps 50% in pivot from search to AI agents

It’s estimated that the AI platform’s annual recurring revenue increased to more than $450 million last month, having launched a new agent tool, and shifted to usage-based pricing, according to the Financial Times.

What we’ve covered

Who is OpenAI’s global head of ads, David Dugan?

We’ve known that CEO of Applications Fidji Simo was looking for an ads boss since last September. Now, David Dugan, has been tasked with that role – Digiday talked to a number of people who know him to understand why OpenAI felt he was the right person for the job.

TikTok’s Khartoon Weiss to exit the platform as its N. America lead of global business

Another top exec exit for TikTok as Khartoon Weiss is leaving the platform today, according to a memo that Digiday has seen, having spent nearly six years in the business.

The Trade Desk is changing how advertisers buy — and what they can see

Within its Kokai platform, The Trade Desk has been testing two new automated buying modes that streamline campaign management, but roll media, data and tech fees into a single price.

What OpenAI’s TBPN deal reveals about branded entertainment’s limits

TBPN’s value is in the fact that Meta CEO Mark Zuckerberg, who rarely does press, has sat down with its hosts. That Microsoft boss Satya Nadella has. That Altman himself keeps coming back.