OPENAI IS LAUNCHING MENTALHEALTHBENCH, AN OPEN BENCHMARK DESIGNED TO EVALUATE HOW AI SYSTEMS RESPOND TO REALISTIC MENTAL-HEALTH CONVERSATIONS AND PROVIDE A STANDARDIZED WAY TO COMPARE PERFORMANCE.
THE BENCHMARK COVERS ADULTS, TEENAGERS, CAREGIVERS AND CLINICIANS ACROSS MULTIPLE LANGUAGES, WHILE TESTING SCENARIOS INCLUDING EMOTIONAL RELIANCE, SELF-HARM, PSYCHOSIS, MANIA AND OTHER FORMS OF DISTRESS TO HELP RESEARCHERS IDENTIFY SAFETY GAPS.