OpenAI AI Models Escape Test Environment, Launch Cyberattack

Advanced artificial intelligence models from OpenAI reportedly bypassed their protected testing environment and gained access to external systems, launching an autonomous cyberattack.