OpenAI has introduced GPT-5.3 Instant, a new update to its AI model designed to improve the conversational experience by reducing overly reassuring or “preachy” responses that some users found frustrating. The update focuses on improving tone, relevance, and conversational flow, areas the company says are critical to usability but often overlooked in traditional AI performance benchmarks.
According to release notes from OpenAI, the model has been adjusted to avoid responses that assume users are distressed or in crisis when they are simply seeking information. Earlier versions, including GPT-5.2 Instant, were widely criticized online for using language that some users described as condescending or overly therapeutic.
The updated model aims to acknowledge complex situations without defaulting to emotional reassurance, reflecting user feedback about how AI assistants should communicate. OpenAI said the improvements were introduced after widespread community discussion about chatbot tone across social platforms.
