One Step Closer to Ads in ChatGPT

It is happening. OpenAI is preparing to introduce ads in ChatGPT. For e-commerce, this marks the beginning of an entirely new advertising format. We are moving away from classic keyword-based search into an environment where customers are solving their problems, asking unanswered questions, and seeking recommendations.

At the same time, OpenAI has officially communicated that it plans to start testing ads in the US, targeting logged-in adult users on the Free and ChatGPT Go tiers. Paid versions – Plus, Pro, Business, and Enterprise – are expected to remain ad-free.

Until now, ChatGPT’s monetisation has been built primarily on subscriptions. With the announced changes, OpenAI is also expanding the lower-cost ChatGPT Go plan at USD 8 per month.

According to OpenAI, the goal of the test is to keep the free version usable.

What Will Advertising in ChatGPT Look Like?

According to OpenAI, the initial format will appear at the bottom of responses when a relevant sponsored product or service emerges within the current conversation. Ads are meant to be clearly labelled and separated from the organic response. Users should be able to see why a particular ad is being shown and have the option to dismiss it.

OpenAI has also outlined several principles. Ads will not interfere with answers. Responses should remain separate and optimised for usefulness, not advertising.

Advertisers will not be given access to user conversations, and user data will not be sold to them. Users will be able to turn off ad personalisation and delete any data used for advertising purposes at any time. Ads will not be shown for sensitive or regulated topics such as health, mental health, or politics. They will also not be displayed to accounts under the age of 18, or where the system assumes this to be the case.

Impact on E-commerce and Marketing

In traditional channels, advertising often works as an interruption, something designed to capture attention and disrupt browsing. In ChatGPT, ads are expected to appear at the moment when a user is actively asking for specific advice. This makes it even more critical whether the ad feels natural and genuinely useful or simply intrusive.

This represents a completely different signal of intent compared to standard search.

Ad Buying and Measurement

Reuters reports that in the pilot phase, OpenAI plans to charge for ads based on impressions rather than clicks. At the same time, according to The Information, the company does not yet have a fully developed system that would allow advertisers to buy ads independently, and is currently working on a self-service solution.

Brand Safety and Display Rules

OpenAI states that ads will not be shown alongside sensitive or regulated topics such as health, mental health, or politics. During testing, ads are also not expected to appear for accounts where the user has stated they are under 18, or where the system infers this.

What the First Format Will Look Like

OpenAI plans to begin by testing ads placed at the bottom of ChatGPT responses, triggered when a related sponsored product or service aligns with the ongoing conversation. Ads should be clearly labelled and separated from the organic response. Users will have the option to see why an ad is being shown, or to dismiss it and provide a reason.

For e-commerce and marketing, these are significant developments. It will now be crucial to monitor how exactly OpenAI defines the initial targeting and exclusion rules, and who will control them. Equally important will be whether impression-based pricing is confirmed and what form reporting takes, so advertisers can realistically evaluate what the ads deliver.

Open questions remain around when campaign management without direct involvement from OpenAI will become available, as well as how quickly ads will expand to other countries and beyond the Free and Go plans.