OpenAI has announced that advertising will soon appear inside ChatGPT conversations, a significant change to one of the world’s most widely used artificial intelligence platforms.

The company said it will begin testing ads for logged-in adult users in the United States who use the free version of ChatGPT, as well as a new lower-priced subscription tier called ChatGPT Go. The move marks OpenAI’s first formal step into advertising within conversational AI and reflects growing pressure to generate revenue at scale.

Ads will appear at the bottom of ChatGPT responses and will be labelled as sponsored content.

OpenAI stated that advertisements will not influence the answers ChatGPT provides and that users will be able to disable ad personalization based on their conversations.

The company also said it will not sell user conversations or personal data to advertisers.

The newly announced ChatGPT Go plan, priced at eight dollars per month, will include ads despite being a paid product, while higher-priced Plus, Pro, Business and Enterprise tiers will remain ad-free.

OpenAI said Go subscribers will receive additional features compared with free users, including longer memory and increased image generation.

Advertising will not be shown to users under eighteen or within conversations involving regulated topics such as health, mental health or politics.

The decision represents a notable reversal in tone from OpenAI leadership.

Chief executive Sam Altman has previously expressed discomfort with advertising in AI products, describing the combination as unsettling, but acknowledged that monetisation pressures could eventually require it.

OpenAI is seeking to support rapidly growing operating costs as it invests heavily in computing infrastructure and expands ChatGPT’s role in shopping, learning and everyday decision-making.

The company already allows users to purchase products through ChatGPT using integrated checkout tools, and advertising is expected to deepen its commercial capabilities.

While OpenAI emphasises that ads will be kept separate from model outputs, the change introduces new questions about how conversational AI platforms balance trust, intimacy and commercial incentives as they become embedded in daily life.