India, Dec. 8 -- OpenAI has quietly switched off a feature in ChatGPT that showed users app suggestions resembling advertisements, after widespread complaints from paying subscribers who said the prompts felt intrusive and misleading. The company maintains that these were not ads, yet acknowledges it mishandled the rollout.

The controversy erupted when several ChatGPT Plus and Enterprise users shared screenshots showing what looked like promotional messages for brands such as Peloton and Target while interacting with the chatbot. Many questioned why a paid service would surface content that appeared to be commercial in nature.

OpenAI responded by insisting that the messages were not paid promotions. According to the company, the suggest...