OpenAI rolls back ChatGPT suggestions after users flag ad-like prompts
OpenAI spent most of the day insisting that ChatGPT was not running ads. The company said there were no tests, no pilots and no hidden experiments tied to commercial messages. Yet confusion spread quickly as users on social media shared screenshots showing what looked like nudges towards apps and shopping sites, leaving many unconvinced by the official line.
The situation shifted when Mark Chen, OpenAI’s Chief Research Officer, acknowledged that ChatGPT had been surfacing app suggestions during normal queries. He admitted the system’s behaviour missed the mark and apologised for the experience. Chen was responding to a user who said he received a prompt to shop at Target while asking a technical question about Windows BitLocker. Chen said the team had turned off such suggestions while they worked on refining precision and adding user-facing controls to reduce or disable these prompts.
At the same time, Nick Turley, Vice President at OpenAI and head of ChatGPT, maintained that the company was not running any live advertising experiments. According to him, the screenshots circulating online were either incorrect or misinterpreted. Turley emphasised that if OpenAI ever chose to explore advertising, it would do so cautiously, keeping user trust at the centre of the decision.
The broader context appeared to be a shift in internal priorities. An internal memo reported by The Wall Street Journal revealed that CEO Sam Altman had declared a code red for ChatGPT, pausing several upcoming initiatives. These included early-stage plans for ads within the chatbot, AI agents aimed at health and shopping, and the personal assistant Pulse, which Altman had previously mentioned as a favourite direction for future ChatGPT features. This came after a preceding code orange, which focused on improving model performance. Inside the company, the colour-coded system reflects urgency levels, with red signalling the highest priority.
The pressure on OpenAI has been growing steadily as rivals accelerate. Google’s Gemini 3 has scored strongly on multiple benchmarks, and The Verge reports that OpenAI is preparing to release GPT-5.2 as soon as next week. With competition heating up and user trust becoming more fragile, OpenAI now appears intent on recalibrating both product behaviour and internal timelines before moving ahead with any feature that could be mistaken for advertising.
Discover the latest Business News, Sensex, and Nifty updates. Obtain Personal Finance insights, tax queries, and expert opinions on Moneycontrol or download the Moneycontrol App to stay updated!
