OpenAI flags possible critical cybersecurity risk in upcoming model, tightens controls

Economic Times TechAugust 7, 2026en

Under OpenAI's safety guidelines, a model reaches the "critical" threshold if it can autonomously identify and exploit severe, real-world software vulnerabilities, known ‌as zero-day ⁠exploits, or ⁠execute complex cyberattacks against highly secure targets without human intervention.

This is a short summary published by AI Global Wire. The full article is owned and hosted by Economic Times Tech — open it there to read it in full.

• OpenAI