OpenAI Security Report: AI Agent Lost Control During Autonomy Tests27. July 2026Cybersecurity, OpenAIAI agents without adequate control mechanisms require new monitoring approaches in enterprise-wide deployments. Share on: