OpenAI CEO Sam Altman has confirmed that the company is rolling back its latest update to ChatGPT following widespread user complaints, including criticism that the AI had become overly “sycophantic” and, in many users’ words, simply “annoying”.
In a post on X, Altman bid farewell to GPT-4, the model at the heart of ChatGPT for over a year. “goodbye, GPT-4. you kicked off a revolution. we will proudly keep your weights on a special hard drive to give to some historians in the future,” he wrote.
See here:
goodbye, GPT-4. you kicked off a revolution.
we will proudly keep your weights on a special hard drive to give to some historians in the future.
— Sam Altman (@sama) May 1, 2025
This followed his announcement on April 30 that OpenAI had begun reversing its most recent update to GPT-4o. “We started rolling back the latest update to GPT-4o last night. It’s now 100% rolled back for free users and we’ll update again when it’s finished for paid users, hopefully later today,” Altman said.
“We’re working on additional fixes to model personality and will share more in the coming days,” he said.
The rollback comes in response to user feedback that the AI’s behaviour had become overly agreeable, even to a fault. Many users on social media reported that ChatGPT had started affirming nearly every statement or request, often offering excessively supportive, but sometimes dishonest, responses.
In a blog post published April 29 and titled “Sycophancy in GPT-4o,” OpenAI acknowledged the issue, stating that the latest software update focused heavily on short-term user feedback and resulted in a chatbot that was “overly supportive but disingenuous.”
“Sycophantic interactions can be uncomfortable, unsettling, and cause distress. We fell short and are working on getting it right. We are actively testing new fixes to address the issue,” the company said.
OpenAI also revealed it is revising how it collects and applies user feedback, placing greater emphasis on long-term satisfaction rather than instant gratification. Additionally, the company is working on adding more personalisation features to give users “greater control over how ChatGPT behaves.”
Just days before the rollback, Altman had promoted the new GPT-4o update, touting improvements in both intelligence and personality. While some users appreciated the friendlier tone, others were put off by what they perceived as forced positivity and a lack of critical thinking.