OpenAI looks at chatbot ads
Add Axios as your preferred source to
see more of our stories on Google.

Illustration: Annelise Capossela/Axios
OpenAI's potential plan to incorporate ads in products like ChatGPT is hardly surprising — but should still serve as a wake-up call for AI makers and users.
Why it matters: AI chatbots cost a fortune to run and remain available for free. Advertising could open up huge new revenue opportunities for OpenAI and its competitors, but it also risks separating the interests of AI companies from those of their customers.
Driving the news: OpenAI CFO Sarah Friar told the Financial Times that OpenAI is weighing the inclusion of ads in its products but wants to be "thoughtful about when and where we implement them."
- Other AI providers have also begun exploring or incorporating ads in chatbots and AI-powered search engines. That includes Microsoft and Perplexity, as well as startups like Adzedek.
Zoom in: There is a range of means by which advertising can merge with the conversational experience of a chatbot — and the "how" is important.
- Currently, creators of custom chatbots can incorporate ads using technology from Microsoft and Adzedek. These tools will deliver a related but distinct and clearly marked ad along with the response to a user's query.
- OpenAI and others could do something similar. As long as the ads are clearly marked and don't otherwise affect a chatbot's response, the experience is comparable to what happens today with search engine ads.
- But if these lines aren't clear, the ads will call into question whose needs a chatbot is really serving — and undermine users' trust in the technology.
The big picture: Advertising has long been the primary revenue source for most consumer internet services, allowing companies to offer everything from email to social media largely for free.
- The approach has proven widely popular, with paid consumer email service relegated to a niche. For-pay social networks, meanwhile, have largely failed to attract sufficient critical mass and flopped, though LinkedIn and X have adopted freemium models.
Yes, but: Ad-based business models have driven internet companies to ratchet up engagement strategies and track users' activities for better ad targeting.
- AI makers will be tempted toward similar tactics — with an additional risk arising from the technology's known power to persuade and bond with users.
- A chatbot-turned-salesman could be both lucrative and annoying. An AI companion that sneaks in a sponsored message, or an AI agent that makes choices for you based on ad dollars, could be predatory.
Between the lines: OpenAI, in particular, has been looking to reassure investors and the broader business community that it has a plan to boost revenue as it continues to rack up giant losses and announce pricey new projects.
- The company raised $6.6 billion in an October funding round, but is also losing billions of dollars per year.
- To date, its revenue has come from a mix of business customers along with consumers who pay $20 a month for a premium version of ChatGPT.
What they're saying: An OpenAI representative told Axios the company has no current plans to implement advertising.
What we're watching: Meta has been aggressively pushing its AI assistant into Facebook, Messenger, Instagram and WhatsApp.
- The social media giant, whose business is fueled almost entirely by advertising, has also said it plans to use AI to generate new content.
Go deeper: I talked more about what ads in chatbots could mean during an appearance on NPR's "Here & Now."
