The AI company acknowledges that last week's update to ChatGPT made it too informal and friendly, prompting it to pull its release.

After a recent update turned ChatGPT into a yes-man, OpenAI rolled it back. But experts say the deeper problem—AI that flatters instead of informs—may be baked into how these systems are trained.

OpenAI has published a postmortem on the recent sycophancy issues with the default AI model powering ChatGPT, GPT-4o.