OpenAI Tests Ads in ChatGPT: What You Need to Know

What’s New

As of February 9, 2026, OpenAI has begun testing advertisements within ChatGPT for a limited number of users in the United States. These ads appear for users on the Free tier and the new low-cost Go subscription ($8/month).

Key details of the rollout:

Placement: Ads appear at the bottom of or visually separate from the chatbot's response, not inside the text answer itself.
Format: They are clearly labeled as "Sponsored" and often include an image and a link to a product or service.
Relevance: Ads are matched to the topic of your current conversation (e.g., asking for a dinner recipe might trigger an ad for a meal kit service).
Exclusions: Paid subscribers on Plus, Pro, Team, and Enterprise plans remain ad-free. Ads are also blocked on sensitive topics like health, mental health, and politics.

Why it matters

This marks a major shift for ChatGPT, which has operated without commercial interruptions since its 2022 debut.

For users: It changes the "clean" utility of the tool. While it keeps the basic version free, users must now choose between seeing ads, paying for a premium subscription, or accepting stricter usage limits to opt out.
For the industry: It signals that subscription revenue alone may not cover the massive costs of running advanced AI models.
Privacy focus: OpenAI has structured this to differ from traditional tracking. They state that advertisers do not receive your chat logs, and the AI model itself does not "see" the ads, maintaining the neutrality of its answers.

Background

Startups like OpenAI face immense operational costs—often billions of dollars annually—to train and run large language models. Until now, OpenAI relied on venture capital and premium subscriptions ($20+ per month) to offset these costs.

Timeline:

November 2022: ChatGPT launches as a free research preview.
February 2023: ChatGPT Plus launches, introducing a paid tier for faster access.
January 2026: Reports surface of OpenAI preparing an ad infrastructure.
February 2026: The "Go" plan is introduced as a mid-tier option, and ad testing officially begins.

Competing platforms have taken different paths. Anthropic (Claude) explicitly markets its interface as ad-free and privacy-focused, while Google (Gemini) has integrated commercial results into its ecosystem for years.

What we don’t know

Conversion rates: It is unclear if users will actually click ads in a conversational interface compared to a traditional search engine.
Global timeline: OpenAI has not confirmed when this test will expand beyond the US or to other languages.
Ad load: We do not know if the frequency of ads will increase over time as the program matures.

What’s next

OpenAI will analyze feedback from this initial US test to refine the experience.

User controls: Users can currently manage ad personalization in their settings or opt out entirely by accepting lower message limits.
Expansion: Expect broader rollouts to other free-tier regions later in 2026 if the revenue model proves successful.
Watch for: New ad formats that might allow users to "ask" the ad questions directly, a feature OpenAI has hinted at for the future.