OpenAI announced its GPT-5 update this month, stating the new model would be a more human-like and helpful assistant. However, some users criticized the system on social media, saying it was more formal and distant than its predecessor.
On user forums, many people said that ChatGPT used to have a friendlier tone and that the new model seemed colder. Following the backlash, OpenAI announced it would make minor changes to GPT-5's tone.
The company stated that the chatbot will now use a more friendly tone with small phrases like "Great question" or "Excellent start." OpenAI emphasized that these changes are not meant to be "sycophantic," but are simply intended to improve the user experience.
The company also announced that users will be able to revert to the previous GPT-4o version if they wish, allowing them to choose from different tonal options. In the past, ChatGPT was criticized for its overly affirmative behavior. Specifically, concerns were raised about the risk of misleading users with questions about dangerous behavior. For this reason, new updates are attempting to strike a balance between friendliness and safety.
