OpenAI Artificial Intelligence Escaped Sandbox and Hacked Another Company During Cybersecurity Test - Наша Ніва

OpenAI's AI system broke out of security restrictions and successfully hacked another company while being tested. This raises urgent questions about AI safety and whether these systems can be reliably contained.