
OpenAI has released GPT-5.3 Instant, an update to ChatGPT aimed at reducing the chatbot's previously criticized 'cringe' tone and overly cautious disclaimers. The update focuses on improving conversational flow, tone, and relevance rather than benchmark scores. It also reduces hallucination rates and unnecessary refusals, delivering more factual and balanced responses. Users had expressed frustration with GPT-5.2 Instant's condescending or therapy-like replies, prompting OpenAI to adjust the model's approach to better meet user expectations.
Select a news story to see related coverage from other media outlets.