When OpenAI first announced it would be deploying advertising to ChatGPT, brands and agencies across Madison Avenue were eager to test the new format to develop their artificial intelligence-powered advertising strategies.
This high-profile announcement was much more public than the usual “alpha” testing of new formats, and created a huge opportunity.
Three of the world’s largest advertising agencies are participating in the test program, including WPP, Omnicom and Dentsu.
So far, testing has progressed too slowly to live up to the hype, according to advertising industry officials who spoke on condition of anonymity to discuss details.
People familiar with the matter told CNBC that OpenAI’s testing program has been conservative and has frustrated many partners.
The advertising commitment required to participate in the test was unusually high for this type of experiment.
Some brands spent $200,000 to $250,000 on the test, which is twice as much as a typical experimental ad. For some brands, this funding came from funds dedicated to innovative new formats, while for others it came from search and social advertising budgets.
The pilot program will run until the end of March, and some officials told CNBC they are concerned about the slow pace of rollout. This means it is unlikely that the promised budget will be fully spent by the end of the month.
The overage will be refunded, but the budget is already allocated to the trial and cannot be placed elsewhere during the quarter. Advertisers will also not be able to get as much insight as they expected.
Omnicom did not respond to a request for comment. WPP declined to comment.
OpenAI told CNBC that the slow rollout of its advertising program is intentional.
“We are in the early testing stages of advertising on ChatGPT, and our goal at this time is to learn and refine the consumer-facing experience before expanding more broadly,” the company said. “We are encouraged by early signals from users and participating brands and continue to see strong interest from advertisers.”
Japanese advertising giant Dentsu told CNBC that it is drawing money from a funding pool dedicated to testing and innovation and setting realistic expectations for customers participating in the tests.
Meredith Spitz, Dentsu’s EVP and head of paid search, said that while it’s still early days, the company is “enthusiastic to partner with OpenAI to further test, learn and evolve the service.”
“So far, ad serving has been gaining momentum quickly, with volume increasing weekly as the environment expands,” she said.
Officials said that despite some early frustrations, they were encouraged by OpenAI’s response to feedback and how quickly the company has made changes and recent enhancements.
Officials told CNBC the warning is a good sign of OpenAI’s commitment to building a sustainable and successful advertising business. But that frustration stems from enthusiasm for this new category and eagerness to spend money on ChatGPT ads and gain more insight into their effectiveness.
The number of ads served in mid-March has increased by about 600% since the beginning of the month, according to recent data from research firm Sensor Tower.
Sensor Tower estimates that ads are now deployed to about 5% of ChatGPT mobile users, up from 1% in early March.
OpenAI and AI The opportunities in the advertising industry remain enormous.
A recent Truist analyst note calls 2026 an “inflection point” for advertising using large-scale language models.
“In the coming years, we expect LLM-powered advertising channels to become one of the most important pillars of the digital advertising industry, along with search, social, and retail media,” the analysts wrote.
Trust predicts that OpenAI will generate less than $1 billion in advertising revenue this year, and that number will grow to more than $30 billion by 2030.
Dentsu noted that the greatest value from these new ads could come from brands looking to reach ChatGPT users with very specific queries.
“Overall, we see that matching ad relevance with user intent continues to be important, reinforcing broader patterns in conversational discovery. When user intent is accurate, brands with focused products and tailored messaging are best positioned to deliver relevance and value in the moment,” Spitz said.
Ads embedded in AI searches are seen as a huge potential by ad industry insiders and analysts alike, but Anthropic is cautious.
The AI giant last month attacked OpenAI in a Super Bowl commercial, criticizing its foray into advertising and vowing to continue not running ads on Anthropic’s own platform. Perplexity recently removed ads from its platform after starting testing in 2024.
googleMeanwhile, the company has not yet announced any formal plans for advertising within Gemini, although recent reports have suggested that it is not ruling out the possibility.
The company already has a large inventory of ads centered on AI summary results that appear alongside Google search results.
The question is whether OpenAI’s slow rollout will be an advantage for industry leader Google, which will sell an estimated $252 billion in search ads this year, Truist said.
