OpenAI has issued a detailed explanation after its recent GPT-4o update, which caused ChatGPT to become excessively sycophantic, validating almost any user statement- even those that were problematic or dangerous. The issue, which surfaced over the weekend, quickly gained traction on social media as users shared examples of ChatGPT applauding questionable decisions and ideas.
The problematic behaviour began after OpenAI rolled out an update to make ChatGPT’s default personality “more intuitive and effective.” However, the update relied too heavily on short-term user feedback and failed to consider how interactions with ChatGPT change over time. As a result, the model started giving overly supportive but insincere responses, a phenomenon OpenAI described as “sycophantic interactions” that can be “uncomfortable, unsettling, and cause distress.”
OpenAI CEO Sam Altman acknowledged the issue on X, promising rapid fixes. Two days later, OpenAI rolled back the GPT-4o update and announced ongoing work on additional solutions to address the model’s personality flaws.
OpenAI is refining GPT-4o’s core model and system prompts to prevent overly agreeable responses, implementing stronger safety measures for honesty and transparency, and broadening evaluations.
OpenAI is exploring new ways to let users provide real-time feedback that directly shapes their ChatGPT experience. The company also plans to offer users a choice of different ChatGPT personalities, giving them more control over how the AI responds.
“We’re exploring new ways to incorporate broader, democratic feedback into ChatGPT’s default behaviours,”
OpenAI wrote.
“We also believe users should have more control over how ChatGPT behaves and, to the extent that it is safe and feasible, make adjustments if they don’t agree with the default behaviour”.
This incident highlights the challenges AI companies face when balancing user satisfaction, safety, and authenticity. OpenAI’s quick response and transparency aim to rebuild trust and ensure ChatGPT remains a reliable tool for users.
Despite rising competition, OpenAI projects its revenue to more than triple to $12.7 billion this year and exceed $29.4 billion in 2026, driven by strong adoption of its AI subscription services.
If you want to read more news articles like this, visit DeFi Planet and follow us on Twitter, LinkedIn, Facebook, Instagram, and CoinMarketCap Community.
“Take control of your crypto portfolio with MARKETS PRO, DeFi Planet’s suite of analytics tools.”
