The dream of serving ads to AI agents has hit a snag

1 week ago 39

Now the bots see them. Now they don't.

Brands including Ally Bank and the Project Management Institute recently began testing ads on the publisher Time's website, specifically designed to be read by AI bots.

The experiment sparked a lively ad industry debate: Could this nascent form of advertising grow into a credible channel, or would the practice be quickly thwarted by the AI platforms they are hoping to target?

AI search company Perplexity decided it doesn't want to play ball.

Digiday reported that Perplexity has begun preventing these ads from influencing its models. In a statement, the company warned publishers that running "deceptive advertising," such as this new format, could result in them being downranked in its search index.

A quick recap:

Time recently created stripped-down, text-only versions of its webpages designed to be read by AI agents.
Time is directing AI crawlers to read those pages rather than the ones humans see.
It then began working with the adtech company Mobian to place "Agent Ads" on those pages.
These ads render as blocks of FAQ-like text with up-to-date facts about the advertised brand.

The central idea is that the ads could help companies leverage the "authority" of publishers like Time and improve their chances of showing up accurately inside chatbots like ChatGPT and Gemini.

Jonah Goodhart, CEO of Mobian, said agent ads are clearly labeled and present cited, verifiable facts about brands that models can choose whether to use.

"This is the opposite of deception," Goodhart told me.

Perplexity spokesman Jesse Dwyer said that serving one page to people and a different page to their software is called "cloaking," which search engines have long penalized.

"You can't just rebrand spam," Dwyer told me. "If the ads are so helpful, why not just put them right on the site for everyone to see?"

I also asked a few other prominent AI platforms whether they would follow Perplexity's lead. OpenAI and Google declined to comment, and Anthropic didn't respond.

Time COO Mark Howard said in a statement that the publisher has contacted Perplexity and is looking forward to "a constructive dialogue about the standards and safeguards that will allow advertising and AI to develop responsibly together."

Meanwhile, PMI CMO Menaka Gopinath described its AI ads experiment with Time as part of its "learning journey" to see how information about project management, skills, and career development surfaces in AI-driven environments.

Like many marketing experiments, sometimes it's not all about the ROI, but the lessons we learn along the way!

Lara O'Reilly is the anchor of the CMO Insider newsletter.She is a senior correspondent who has covered the digital advertising, marketing, and media industries since 2010. Her current beat includes big tech companies like Alphabet, and Meta, and adtech firms, agencies, publishers, the creator economy, and CMOs.Lara has previously worked as a reporter and executive producer at titles including The Wall Street Journal, Digiday, Yahoo Finance, and Marketing Week. She was previously Business Insider's senior global advertising editor from 2014 to 2017.Lara was named "Digital Journalist of the Year" by the London Press Club in 2016.Lara is a regular guest on TV and radio and has appeared on outlets such as the BBC, NPR, SiriusXM's Wharton Business Daily, and CTV Television Network. She also frequently speaks on stage at major events such as Web Summit, IFA, VivaTech, Advertising Week, and Cannes Lions.To get in touch with Lara O'Reilly, email [email protected] or contact her on Signal at @loreilly.71Check out Insider's source guide for tips on sharing information securely.Read some of Lara's recent work below: