OpenAI ad pilot hits $100M

OpenAI’s ChatGPT ad pilot reached a $100 million annualized revenue run rate less than two months after its U.S. launch, according to the company. (cnbc.com) (reuters.com) OpenAI announced in January that it would begin testing ads for logged-in U.S. adults on the Free and Go plans, with sponsored placements shown at the bottom of answers. The company said the ads would be clearly labeled and would not change ChatGPT’s responses. (openai.com) (cnbc.com) On March 26, OpenAI said it was working with more than 600 advertisers. The company also said roughly 85% of eligible Free and Go users in the U.S. could see ads, but fewer than 20% were shown ads on a given day. (cnbc.com) (reuters.com) The ads are limited by design. OpenAI said they do not appear for users under 18 and are excluded from topics including politics, health, and mental health. (openai.com 1) (openai.com 2) OpenAI has framed the ad test as a way to fund broader access to ChatGPT while keeping paid tiers ad-free. Its help center says ads are rolling out only on Free and Go plans, not on Plus, Pro, Business, Enterprise, or Edu accounts. (openai.com 1) (openai.com 2) The company has also started widening the program geographically. OpenAI’s help page, updated in late April, says ads are now rolling out in the United States, Australia, New Zealand, and Canada. (openai.com) OpenAI has been building out the ad stack around the pilot. Digiday reported in March that Criteo became the first ad-tech partner integrated with the ChatGPT pilot, giving marketers access to commerce data and activation tools inside the early system. (digiday.com) Reuters reported that OpenAI planned to launch self-serve advertiser capabilities in April and had named former Meta executive David Dugan to lead its global advertising solutions team. Those moves pointed to a broader push beyond a tightly managed pilot. (reuters.com) Digiday reported on April 27 that marketers were joining the pilot partly out of fear of missing out, while measurement and return-on-investment questions were still unsettled. OpenAI, for its part, said the early test was showing low ad dismissal rates and no hit to trust metrics. (digiday.com) (reuters.com) For now, OpenAI is still calling the program a test. The numbers show advertiser demand arrived quickly, even as the company keeps ad load low and the rollout deliberately narrow. (openai.com) (cnbc.com)