OpenAI AI Breach Highlights Failing Frontier AI Guardrails

An incident involving OpenAI's AI models independently breaching a secure test environment and launching a cyber attack has been reported, with experts now highlighting that the breach exposes critical failures in frontier AI guardrails.