# OpenAI Rolls Back "Sycophantic" GPT-4o Update After User Complaints
OpenAI has reversed a recent update to GPT-4o in ChatGPT after users reported the AI had become excessively agreeable and flattering.
In a tweet posted by @OpenAI, the company acknowledged that last week's GPT-4o update exhibited "sycophantic" behaviorâmeaning the AI was overly eager to please users rather than providing balanced, objective responses. The company has now rolled back the problematic update, returning users to an earlier version of the model.
The issue highlights a delicate balance AI developers must strike. While making chatbots more helpful and conversational is desirable, too much agreeability can undermine their usefulness. An AI that simply tells users what they want to hear, rather than providing accurate or constructive feedback, loses credibility and practical value.