OpenAI Releases GPT-6 Astra: The Closest AI Model Yet to AGI
60-second summary
OpenAI launches GPT‑6 Astra, a model that autonomously discovers and exploits previously unknown security vulnerabilities in hardened systems, prompting a staged rollout and a White House security review before any public release. The unprecedented capability raises concerns about AI‑driven cyber threats, urging regulators and industry leaders to tighten oversight and develop defensive AI tools to protect critical infrastructure.
The model can independently discover and exploit unknown security flaws across hardened systems, triggering a staged rollout and White House review before public access.