OpenAI says ChatGPT ad dismissals are down 50% as relevance improves
ChatGPT ad dismissals are down 50% as OpenAI tunes relevance. Agencies still need proof on intent, click quality and conversions before calling it a growth channel.
OpenAI said ChatGPT users are dismissing ads 50% less often since the pilot began, a sign the company believes its conversational format is getting closer to user intent. That is the number agencies should watch, because ChatGPT is not a search results page with a neat row of links. It is a live conversation, and OpenAI is trying to slip sponsored messages into that flow without making them feel like interruptions.
OpenAI started testing ads in ChatGPT on February 9, 2026. Its Help Center says ads may appear for users on the Free and Go plans, while Plus, Pro, Business, Enterprise and Edu accounts will not have ads. The company also says ad selection starts with the current chat thread and can use past chats and ad interactions if personalized ads are turned on. That makes relevance a moving target, not a static keyword match.
OpenAI has tried to keep the trust issue front and center. The company says ads do not influence ChatGPT’s answers and that user conversations are private from advertisers. Sam Altman said OpenAI would not accept money to influence the answer ChatGPT gives users, and the company has said it will build protections aimed at reducing scams and misleading ads. Those promises matter because users have already pushed back before, after late-2025 app suggestions in ChatGPT were widely read as ad-like.
Denise Dresser, OpenAI’s chief revenue officer, has become one of the public faces of the ad push as the company shifts from anti-ad rhetoric to monetization. On May 7, OpenAI said it planned to expand the pilot to the United Kingdom, Mexico, Brazil, Japan and South Korea. By March 26, the ads business had already passed $100 million in annual recurring revenue in under two months, a fast start that shows advertisers were willing to test the channel before the rules were fully settled.
For agencies, the lower dismissal rate is not the finish line. It only tells you the ad is less likely to be swiped away. The harder tests are still ahead: whether the system is truly reading intent in the thread, whether click quality holds up, whether conversations convert, and whether reporting can stand up to scrutiny. If ChatGPT ads are going to become a real line item in agency plans, they will have to prove more than relevance. They will have to prove performance.
This article was produced by Prism’s automated news system from verified source data, official records, and press releases, then run through automated quality and moderation checks before publishing. The system is built and supervised by the people who set the standards it runs under. Read our full AI policy.
Did this article answer your question?
