OpenAI announced an update to its GPT-5 model late Friday, intending to make the artificial intelligence “warmer and friendlier” following user feedback indicating a preference for the prior GPT-4o model.
The company launched GPT-5, a process CEO Sam Altman described as “a little more bumpy than we’d hoped for.” This update aims to address user complaints about GPT-5’s initial personality, with OpenAI stating the changes are “subtle” but will make the model “more approachable now.”
OpenAI elaborated on the nature of these changes in a social media post, noting, “You’ll notice small, genuine touches like ‘Good question’ or ‘Great start,’ not flattery.” The company also reported that “Internal tests show no rise in sycophancy compared to the previous GPT-5 personality.”
We’re making GPT-5 warmer and friendlier based on feedback that it felt too formal before. Changes are subtle, but ChatGPT should feel more approachable now.
You'll notice small, genuine touches like “Good question” or “Great start,” not flattery. Internal tests show no rise in…
— OpenAI (@OpenAI) August 15, 2025
During a recent dinner with journalists, OpenAI executives, including VP Nick Turley, discussed future plans. Turley stated that the initial GPT-5 was “just very to the point,” acknowledging the model’s directness before this announced update to enhance its warmth.
