On February 9, 2026, OpenAI crossed a line it had long resisted. After years of positioning ChatGPT as a neutral, commercial-free tool, the company began serving advertisements to a subset of its users — marking the formal start of what many in the industry are calling the Generative Engine Advertising era.
Here is what we know, based on OpenAI’s official announcement and early reporting.
Who sees the ads?
Advertisements are currently shown only to users on the Free and Go tiers ($8/month). Users on Plus ($20/month), Pro ($200/month), Team, Enterprise, and Education plans will not see ads. OpenAI is framing this as a trade-off: free access in exchange for a commercial experience, while paid subscribers retain an ad-free product.
The initial rollout is limited to logged-in users in the United States. Expansion to other markets is expected throughout 2026, though no specific timeline has been confirmed.
What do the ads look like?
Two formats have been confirmed in the beta:
- Sponsored product cards — when a user asks a question related to food, shopping, or products, a card appears below the AI’s response showing a brand logo, product name, price, stock status, and estimated delivery or preparation time. Clearly labelled as “Sponsored”.
- Contextual text placements — text-based sponsored content appearing below the organic response, matched to the intent of the conversation.
Crucially, OpenAI has formalised an “Answer Independence” principle: advertisements never influence what ChatGPT says. The model generates its response independently of any advertising considerations, and ads are served separately — after the answer has been generated.
How much does it cost to advertise?
Access to the beta comes at significant cost. OpenAI is charging approximately $60 per thousand impressions (CPM) — roughly three times Meta’s average rates and significantly above most programmatic display advertising. The minimum commitment to participate in the beta is $200,000.
That entry bar puts the current programme out of reach for the vast majority of advertisers. For now, it is a closed experiment with a small group of approved brand partners.
What OpenAI says about privacy
OpenAI has stated that advertisers receive no access to individual user conversations, chat history, memories, or personally identifying information. Targeting is based on the content of the conversation in aggregate — the topic being discussed — not on individual user profiles.
What this means
The ChatGPT ad launch is significant not because of its immediate scale — access is limited and expensive — but because of what it signals. OpenAI has 400 million weekly active users. Even with ads restricted to Free and Go tiers, the potential reach is enormous. The February 2026 beta is a proof of concept. If it performs, broader access — and potentially lower minimum commitments — will follow.
For advertisers watching from the sidelines: the question is not whether GEA will become a mainstream channel, but when. OpenAI has indicated plans to introduce more ad formats and expand access throughout 2026.
Sources: OpenAI — Our approach to advertising · TechCrunch — ChatGPT rolls out ads · ALM Corp — ChatGPT Ads analysis