OpenAI has released a system card addendum for GPT-5.1 Instant and GPT-5.1 Thinking models, detailing updated safety metrics. This document is crucial for developers and AI safety researchers as it introduces new evaluations specifically targeting mental health scenarios and emotional reliance. By providing transparent data on how these advanced models handle sensitive user interactions, OpenAI aims to foster responsible AI deployment.

These updates highlight the industry's growing focus on mitigating psychological risks associated with conversational AI, ensuring that models like GPT-5.1 are equipped to support users without causing harm. This move reinforces the importance of ethical guidelines in the rapidly evolving landscape of generative AI.