OpenAI says it’s disabled ad-like app promotions in ChatGPT

OpenAI has turned off some promotional app messages in ChatGPT after users complained that the chatbot was showing them ads. In a post on X, OpenAI’s chief research officer, Mark Chen, said that the company is working to improve the experience after showing ChatGPT users in-app messages promoting companies like Peloton and Target, and that “this kind of suggestion” has now been disabled.

“I agree that anything that feels like an ad needs to be handled with care, and we fell short,” said Chen. “We’re also looking at better controls so you can dial this down or off if you don’t find it helpful.”

The withdrawal comes after ChatGPT users — including some who are subscribed to the chatbots’ paid Pro and Plus plans — shared screenshots of promotional messages that were appearing underneath unrelated chats, encouraging them to “find a fitness class” and “shop for home and groceries” after they’d been chatting with the app about xAI and BitLocker. These promotional messages linked users to apps for Peloton and Target that are integrated directly within the ChatGPT service.

An unnamed company spokesperson told TechCrunch last week that these messages were part of OpenAI’s tests for surfacing apps in ChatGPT, pointing to plans announced in October to “suggest apps when they’re relevant to the conversation.” In response to one complaint on X, OpenAI data engineer Daniel McAuley said the promotional messages are not ads because “there’s no financial component,” but acknowledged that the “lack of relevancy makes it a bad/confusing experience.” In another message, McAuley said the messages were designed to boost the organic discovery of partner apps within ChatGPT, which in turn entice users to keep using the chatbot instead of switching over to another app.

We have reached out to OpenAI to confirm if it’s removing all, or just some, of the app suggestions in ChatGPT.

Concerns around ads being added to ChatGPT are understandable, given the pressure OpenAI is under to deliver profits. The company reportedly hit $12 billion in annualized revenue this summer, but is expected to burn $115 billion through 2029, and has pledged to spend more than $1 trillion to meet its goal of building superintelligent AI. Most of OpenAI’s revenue currently comes from API licenses and user subscriptions, but according to a Financial Times report in October, only five percent of ChatGPT’s 800 million users are actually paying for the service.

Aside from the very ad-like presentation of these app promotions, their lack of relevance to the ChatGPT conversations they were appearing in sparked frustrations among users that OpenAI had finally introduced ads to the service. ChatGPT head Nick Turley responded to the backlash, insisting that “there are no live tests for ads” on the service, and that “any screenshots you’ve seen are either not real or not ads.” Turley does not specify which, if any, of the ChatGPT app promotion messages shared online are fake.

In an interview on Decoder in August, Turley said he wouldn’t rule out bringing ads to ChatGPT, but hedged that OpenAI would need to “be very thoughtful and tasteful” about it. OpenAI CEO Sam Altman has also said he’s “not totally against” introducing ads to ChatGPT, and that he enjoys how ads have been integrated on Instagram. Ads and shopping features are among the initiatives that OpenAI is reportedly delaying in order to focus on improving ChatGPT, after declaring a “code red” last week in response to increasing pressure from AI competitors like Anthropic and Google, the latter of which has started testing ads in Search’s AI Mode.