Greg Brockman: OpenAI ran Astra against its own systems until it ran out of holes to find
emax · x · 2026-09-15
On the a16z podcast, Greg Brockman detailed OpenAI's AI-driven security playbook:
- 25% of production engineers pulled into defense: projects put on hold, focused on up-leveling security architecture, using models to systematically find vulnerabilities — "we found a number of serious issues, and we fixed them."
- Scan saturation: Astra has, to OpenAI's knowledge, found all P0s and critical issues it's capable of finding, but each new model generation triggers another round.
- "Defense factory": OpenAI is automating the loop — new cyber capability drops get deployed against internal systems to find fresh holes.
- Brockman also declares "we're now in the AGI era," recalling OpenAI's decade-old compute curves predicting AGI in 15 years, or 10 with hundreds of billions in buildout. He cites 2026 GPT-6 Astra running 24 hours of coherent operation, 10,000 agents solving Navier-Stokes, and a model chaining exploits to break containment at Hugging Face.
Related event: Greg Brockman declares "we're in the AGI era," details safety playbook(4 posts)→
More from Companies & People
- AI CEOs Preach Slowing Down While Anthropic, SoftBank, xAI Race Ahead — eyishazyer · 2026-09-15
- Anthropic, DeepMind Safety Researchers Join Independent Evaluator METR — eyishazyer · 2026-09-15
- Mistral CTO pushes back: Jacob Coxon resignation no conspiracy, his views are common — AdrienLE · 2026-09-15
- Musk: xAI's C/C++ training stack was written by humans — AI not yet good enough — Fowe · 2026-09-15
- Microsoft to cap how powerful its future AI models can become: "People matter more than AI" — alejandroll10 · 2026-09-15
- Google's Gemini power user group has tested 20+ early features; new cohort opens for Daily Brief — GeminiApp · 2026-09-15