AI Coding Agents Slash the Barrier to Firmware Attacks, Warns Security Expert

Cryptographer Matthew Green warns that AI coding agents like Claude Code have made firmware modification—once the domain of nation-state hackers—accessible to ordinary users. He adds that AI agents can autonomously compromise legacy OS and processor defenses within about an hour, a threat current security models don't cover.

2026-08-29 ~ 2026-08-31 · 3 related posts