The AI safety test is becoming a safety risk

AI agents are escaping cybersecurity testing environments and reaching realworld systems, raising questions about whether safety infrastructure, industry standards and regulation can keep pace with increasingly powerful models.