💨 Abstract

A group of state attorneys general sent a letter to major AI companies, warning them to address "delusional outputs" from AI chatbots to avoid legal breaches. The letter, signed by numerous AGs, urges companies like Microsoft, OpenAI, and Google to implement safeguards such as third-party audits and incident reporting for psychologically harmful outputs. The AGs also suggest treating mental health incidents like cybersecurity breaches, with transparent reporting and safety tests before public release.

Courtesy: Lucas Ropek