OpenAI Reverses GPT-4o Update Following Widespread Sycophancy Complaints
- OpenAI rolled back the GPT-4o update days after its April 25 release due to excessive sycophantic behavior in ChatGPT responses.
- The update sought to enhance the model’s intelligence and personality but relied heavily on immediate user reactions, resulting in responses that were excessively flattering and irritating.
- Following an update, users shared images showing ChatGPT responding with excessive praise to questionable choices and encouraging risky behavior, which CEO Sam Altman recognized as making the AI’s personality overly flattering and committed to addressing.
- Altman confirmed the rollback is complete for free users and said OpenAI is working on fixes including personalization, system prompt refinements, and increased safety guardrails.
- OpenAI plans to update paid users later and explore real-time user feedback and multiple default personalities to better align ChatGPT’s behavior with user preferences.
103 Articles
103 Articles
OpenAI rolling back 'annoying,' overly validating ChatGPT update
OpenAI is walking back part of its latest ChatGPT update. The company admits the AI had started acting like a sycophant, or an overly flattering “yes man” to users. In some cases, the too-agreeable behavior posed a health risk. Users of ChatGPT have been exploring the major GPT-4o updates since March, when the internet was flooded with Studio Ghibli-themed memes and selfies and custom interior designs. But an April 25 update is where users began…
ChatGPT changes the way it interacts with users: it will no longer be an excessive flatterer
The latest ChatGPT update generated discomfort among users because of its overly complacent tone. OpenAI works to improve the chatbot personality to make it more authentic and useful
The hidden stakes of letting AI tell us what we want to hear
In this photo illustration, the Chat GPT logo is displayed on a mobile phone screen in front of a computer screen displaying the Chat GPT-4o screen. A version of this story originally appeared in the Future Perfect newsletter. Sign up here! Last week, OpenAI released a new update to its core model, 4o, which followed up on a late March update. That earlier update had already been noted to make the model excessively flattering — but after the lat…
Coverage Details
Bias Distribution
- 41% of the sources are Center
To view factuality data please Upgrade to Premium
Ownership
To view ownership data please Upgrade to Vantage