OpenAI flags possible critical cybersecurity risk in upcoming model, tightens controls

Full body text is not yet available for this story. Tap “View original source” below to read it on the publisher’s site.