OpenAI's 250-person security sprint: Codex agents wrote every patch across hundreds of systems
rohanpaul_ai · x · 2026-09-10
OpenAI published a case study and reference architecture called the "Defense Factory", describing an internal code-red sprint that mobilized 250+ people, where Codex and its cyber models wrote every patch across hundreds of systems, finding and fixing vulnerabilities humans might never have discovered.
The core argument: attackers can now run fleets of long-running agents on open-weight models, so defenders should convert security work into a continuous agent loop too — agents find vulnerabilities, validate them, and verify fixes. OpenAI shared the architecture and a practical playbook for building your own.
More from coding & agent
- For OSS apps, proofs of concept — not PRs to one repo — may be the core primitive — dbreunig · 2026-10-27
- Hook everything into the Dot: a workaround to stretch ChatGPT subscription limits — teortaxesTex · 2026-10-27
- Jev launches people-search API for agents at $0.0089/lead, claiming 85% cheaper than Clay — alexcovo_eth · 2026-10-27
- Going through every compound engineering skill to decide where humans stay in the loop — kieranklaassen · 2026-10-27
- UCSB CS Prof. Xifeng Wang lands 3 grants worth $180K+ for practical agentic AI research — xwang_lk · 2026-10-27
- Gen games don't need to be realtime — turn-based interactivity works, devs argue — djcows · 2026-10-27