Dive Brief:
- OpenAI will begin testing advertising on ChatGPT, its generative artificial intelligence platform, in the coming weeks, the company announced in a blog post.
- The ads will appear for logged-in users on ChatGPT’s free level, as well as on ChatGPT Go, an $8-per-month option that has been available in some markets and is now rolling out to the U.S. and other areas. ChatGPT’s Plus, Pro, Business and Enterprise subscription levels will not carry ads.
- OpenAI said it would implement several protections to assure users that the information they reveal on the AI platform would not be shared with advertisers and that what ChatGPT provides would be “driven by what’s objectively useful,” rather than advertising.
Dive Insight:
While OpenAI CEO Sam Altman had previously expressed resistance to advertising on the platform, the company needs to raise trillions of dollars to make the investments necessary to remain a leader in an increasingly competitive field. Google, meanwhile, is beginning to catch-up — and perhaps even surpass — OpenAI in the generative AI space with the release of Gemini 3. Google’s latest product is “widely regarded as the best large-language model on the market,” winning in most benchmark tests and topping the list for many tasks, according to The Verge.
The expansion into ads arrives at a time when advertisers continue to view AI positively as a tool for supporting their advertising efforts. However, recent research from the Interactive Advertising Bureau found a widening gap between advertisers’ bullish attitude toward AI in advertising and consumers’ increasingly negative view.
As a result of consumers’ widespread distrust of AI in ads, OpenAI may find itself challenged in convincing the public that it can live up to its promises of creating a generative AI platform that “benefits all of humanity” while also offering advertising. One potential issue is that users may lose faith in the platform if they feel the information being served to them is being influenced by advertising.
To that end, OpenAI said that advertising will not influence the answers ChatGPT provides and that ads and sponsored content would be clearly labeled. At least initially, the ads will appear at the bottom of ChatGPT’s answers when, according to the blog post, there’s a relevant and sponsored product or service based on a user’s conversation. Certain topics, including health, mental health and politics, would not be eligible for advertising, and the platform would not serve ads to users who have either identified as minors or whom the platform believes are under 18 using its age-prediction models.
OpenAI also vowed to keep users’ conversations private, instead only offering aggregated data to advertisers about ad performance, like how many times an ad was clicked on or served across the platform. The company said it would offer users transparency into why they were seeing a certain ad, give them the opportunity to dismiss any ad and enable users the ability to turn off personalization and clear the data used for ads.
OpenAI said it will prioritize user trust and the user experience while insisting that ads can provide additional value for users.
Using ChatGPT’s ads, for instance, users might ask questions about traveling to a city and then see ads about booking hotels or travel for the destination. They can also directly ask the chatbot questions about the advertised products to better inform purchase decisions.
