OpenAI starts testing ads within ChatGPT

OpenAI will begin testing ads on ChatGPT in the US in the coming weeks. Ads will appear at the bottom of chatbot responses, clearly labeled, and will only appear if there is a related sponsored product or service tied to the conversation.

Who will see the ads:

• Older users who have entered the free tier
• Users are on ChatGPT Go, OpenAI’s low-cost subscription of $8/month
• The Pro, Business, and Enterprise plans will remain ad-free
• Users under the age of 18 will not see ads

Why do we care. Ads within ChatGPT open up a new, highly targeted environment where users ask questions and make decisions. Unlike traditional search or social ads, this placement comes directly from relevant conversations, providing context-driven exposure with clear user controls. If scaled, this could be an entirely new channel for engagement and acquisition – especially for brands focused on targeting, education, or consideration stage marketing.

A great goal. OpenAI says the ads are part of a broader plan to make powerful AI accessible to more people. ChatGPT Go, launched in 171 countries last year, is now available in the US, offering expanded features such as image processing, file uploading, and storage at a lower price. Ads are intended to help reduce usage restrictions for free users and keep costs down.

What will not change. OpenAI emphasizes that ads will not influence ChatGPT responses and that user conversations will never be sold to advertisers. Ads will not appear near sensitive or controlled topics such as politics, health, or mental health.

OpenAI Ad Principles. The company says its approach is guided by machine alignment, response autonomy, chat privacy, and user control. Users will be able to see why they are being shown an ad, dismiss it, turn off personalization, or delete ad-related data at any time.

Big picture. With an infrastructure commitment of over $1.4 billion and a reported revenue of $20 billion, OpenAI is under pressure to find sustainable revenue streams. Advertising — long a revenue engine for companies like Google and Meta — can help reduce those costs while AI becomes more widely available.

Between the lines. CEO Sam Altman has previously warned that ads can destroy trust, but OpenAI is putting this test as a cautionary and user-first measure, stressing that it won’t work for wasted time or revenue for the loss of information.

What’s next. Ads are not live yet, but testing will begin soon. OpenAI says it will adjust how ads appear based on user feedback as it tries to monetize its flagship product.