OpenAI begins testing ads in ChatGPT

OpenAI has begun rolling out ads inside ChatGPT, marking a major shift for a product that has largely operated without traditional advertising since its launch in 2022.

In a blog post published this week, the company confirmed it is testing ads for logged-in users on its Free and Go plans in the U.S., while keeping paid tiers like Pro, Business, Enterprise, and Education ad-free. OpenAI said the move will help fund broader access to advanced AI tools without requiring every user to pay a subscription.

"Our focus with this test is learning," OpenAI's blog post read. "We’re paying close attention to feedback so we can make sure ads feel useful and fit naturally into the ChatGPT experience before expanding."

The ads appear outside of ChatGPT’s responses and are clearly labeled as sponsored content. OpenAI says ads do not influence how the chatbot answers questions and that user conversations are not shared with advertisers. Instead, ads are selected based on broad conversation topics and how users interact with ads, with restrictions in place to prevent sponsored content from appearing alongside sensitive topics such as health, mental health, or politics.

Those who use ChatGPT's free service can opt out of the ads, with a caveat.

"If you prefer not to see ads, you can upgrade to our Plus or Pro plans, or opt out of ads in the Free tier in exchange for fewer daily free messages," according to the company.

The opt-out options on ChatGPT found in the user settings Credit: Mashable screengrab via OpenAI

Users who do consent to ads will also have the option to opt out of ad personalization, limiting how sponsored content is selected. There are also options to stop ChatGPT from utilizing past AI chats to tailor ads, as well as deleting "all ads history and data" the company has compiled on a user.

At the time of publication, Mashable attempted to surface ads during regular use of ChatGPT but were unable to trigger any sponsored content, which aligns with OpenAI’s description of the rollout as a limited test rather than a full launch.

The rollout follows months of user confusion and frustration after widely circulated screenshots appeared to show promotional content embedded in ChatGPT responses. OpenAI previously dismissed those incidents as poorly timed "suggestions," but the distinction did little to calm concerns. As Mashable reported earlier this year, OpenAI has been quietly experimenting with ad formats internally while signaling that monetization would eventually be necessary to support the platform’s massive infrastructure costs.

With ChatGPT now testing ads and offering opt-out controls, OpenAI appears to be betting that transparency and choice will soften the transition to a more familiar, ad-supported internet model. However, that shift hasn’t gone unnoticed by competitors. Anthropic, one of OpenAI’s biggest rivals, used its Super Bowl LX ad buys to openly mock the idea of advertising inside AI chatbots.

The ads promote Anthropic’s chatbot, Claude, by staging scenarios where seemingly helpful conversations suddenly pivot into awkward sales pitches, ending with the tagline, "Ads are coming to AI. But not to Claude."

Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.