OpenAI is pulling back on promotional-style suggestions in ChatGPT after paying subscribers complained they resembled ads for brands like Peloton and Target. The company continues to insist there are no ads — or ad tests — currently running in ChatGPT. Still, Chief Research Officer Mark Chen acknowledged the experience “felt like advertising” and said OpenAI “fell short” of user expectations. Those suggestions have now been disabled while engineers refine the model and add user controls to reduce commercial-feeling responses.
Executives clarified the tests were meant to highlight third-party apps built on the ChatGPT platform announced in October, not paid advertisers. But the backlash underscores growing tension as ChatGPT becomes an app ecosystem and commercial product relied upon by millions.
The timing is notable: CEO Sam Altman issued a “code red” directive prioritizing improvements to ChatGPT quality and reliability, delaying new monetization streams such as an anticipated advertising business led by Applications CEO Fidji Simo.
The company is also deepening its AI infrastructure ambitions — including a partnership with Foxconn on U.S. manufacturing of next-generation server hardware — while navigating broader industry debates about licensing and responsible AI use, highlighted by ongoing legal disputes between news publishers and AI platforms such as Perplexity.
