OpenAI announces advertising in ChatGPT

OpenAI has announced that it will display advertisements in its AI chatbot ChatGPT. Initial experiments with this new form of monetization are limited to the US market. The company plans to test advertising in the free version of ChatGPT and in the cheapest paid version, "ChatGPT Go," in the coming weeks. The more expensive subscription variants Plus, Pro, Business, and Enterprise will – at least for the time being – not display advertisements.

ChatGPT Go was originally launched in India in August 2025, gradually expanded to other countries, and has only been available worldwide, including in the USA, since January 16, 2026. On the same day, Fidji Simo, CEO of Applications at OpenAI, announced the advertising experiments in the company blog.

Urgent Problems

The idea of financing through advertising represents a significant change in direction for OpenAI. As recently as October 2024, Sam Altman, CEO of OpenAI, began his answer to a question to that effect by stating that he personally hated advertising. He found a combination of advertising and AI particularly disturbing, for example, when users have to consider who is paying how much for them to see a particular answer. While he did not explicitly rule out OpenAI showing ads in the future, he called it a "last resort as a financing model."

This last resort is now to be pursued, and there have been indications of this for some time. When Altman declared "Red Alert" at OpenAI in early December 2025 due to growing competitive pressure, it was stated that side projects such as work on advertising should be postponed. However, this postponement apparently did not last long, possibly due to the massive losses OpenAI is incurring.

Advertising, but with Principles

In the blog post on the advertising experiments, Simo emphasizes that they will continue to prioritize user wishes and maintain their trust. Advertising in ChatGPT should therefore be guided by five principles. Among other things, this includes that advertising will never influence the language model's responses. Furthermore, advertising should be displayed separately from the response and clearly marked. In terms of content, however, advertising can be aligned with the response and the ChatGPT user. An example image in the blog shows advertising placed below the response, with which users can interact directly.

OpenAI also promises never to share conversation content with advertisers or sell user data to them. ChatGPT users will also be able to turn off ad personalization. OpenAI does not disclose what data is used for personalization but promises that users can delete this data at any time. In addition, there will always be an option to not be shown ads, such as a paid subscription option that does not contain ads.

At least during the current testing phase, advertising will not be shown to minors or appear in sensitive contexts such as (mental) health or politics. OpenAI looks forward to user feedback on this experiment, which will be used to optimize how exactly advertising appears in ChatGPT.

(syt)