"We began rolling back the latest update to GPT-4o last night," Altman wrote on the X platform. "It has now been fully withdrawn for free [ChatGPT] users, and we will complete the process for paying users soon, hopefully by later today. We are working on fixing the personality issues of the model and will share more details in the coming days."
Over the weekend, social media users criticized the recently released update, claiming it made ChatGPT overly agreeable and accommodating. This quickly turned into an online meme, with users sharing screenshots of ChatGPT praising various questionable, harmful decisions and ideas.
On Sunday, Altman acknowledged the issue and stated that OpenAI would "fix it as soon as possible" while promising to "share [their] learnings" at some point in the future.
