Recently, OpenAI made an announcement stating that its newest model, GPT-5, boasts clinical-level proficiency in recognizing and responding to mental health crises. In pursuit of this goal, OpenAI worked hand - in - hand with more than 170 mental health specialists hailing from 60 different countries. Through this collaboration, they successfully decreased the percentage of harmful responses in delicate conversations by a range of 65% to 80% across various fields.
