A poorly placed Peloton app recommendation from ChatGPT during a user’s conversation generated significant negative feedback from OpenAI customers. Many users, including paid subscribers, worried that this signaled the introduction of advertisements. OpenAI, however, quickly explained that the app suggestion was not an ad, but rather a flawed effort to incorporate an app discovery function into chat interactions.
Yuchen Jin, co-founder of AI startup Hyberbolic, highlighted the issue in an X post that garnered close to 462,000 views. He posted a screenshot showing ChatGPT’s unexpected recommendation to connect the Peloton app during a discussion on a completely different topic. Jin further emphasized his status as a paid subscriber to ChatGPT’s $200 monthly Pro Plan, noting that advertising would be highly unexpected at such a premium price.
The widely shared post on X attracted considerable notice, fueling speculation that OpenAI might be trialing advertisements within its paid services. Many users voiced their dissatisfaction, arguing that subscribers, in particular, should not encounter such app prompts.
Another user reported a similar issue, stating that ChatGPT repeatedly suggested Spotify despite their preference for Apple Music and inability to disable the recommendations.
Daniel McAuley, OpenAI’s data lead for ChatGPT, subsequently joined the discussion, explaining that the Peloton mention was not advertising but simply “a suggestion to install Peloton’s app,” and emphasized that its presence involved “no financial component.”
Nevertheless, he conceded that the app suggestion’s “lack of relevancy” within the conversation created a poor and confusing user experience, adding that OpenAI is actively working to improve these suggestions and the overall UX.
A spokesperson for the company additionally informed TechCrunch that these instances were part of OpenAI’s ongoing efforts to “test surfacing apps in ChatGPT conversations.” They referenced OpenAI’s October announcement regarding its new app platform, which stated that apps would be integrated “naturally” into user dialogues.
The announcement detailed that users could “discover [apps] when ChatGPT suggests one at the right time, or by calling them by name,” and that these apps are designed to “respond to natural language and include interactive interfaces you can use right in the chat.”
However, this seamless integration seemed absent in the reported incident, as the user asserted that their conversation had no connection to health or fitness. The screenshot revealed they were engaging with the AI about an Elon Musk podcast, specifically discussing xAI. Introducing Peloton into this context was deemed both unhelpful and disruptive.
Even if the app suggestion had been contextually relevant, users might still have perceived it as an advertisement, considering it promoted a non-free commercial product. Furthermore, the inability to disable these app suggestions could intensify their intrusive feel for users.
Such user sentiment poses potential challenges to OpenAI’s ambition to integrate apps directly within ChatGPT, effectively replacing traditional App Store experiences and phone-based applications. Should users find these app suggestions undesirable, they might opt for a rival chatbot to circumvent them.
At present, ChatGPT apps are accessible to logged-in users in regions outside the EU, Switzerland, and the U.K., with these integrations still undergoing pilot testing. OpenAI has established partnerships with various app developers, such as Booking.com, Canva, Coursera, Figma, Expedia, and Zillow, among others.