Simon Willison: Claude Mythos Created 181 Working Exploits vs Opus 4.6's Near-Zero — 'Industry-Wide Reckoning'

Simon Willison·medium signal

Willison's technical analysis of the Mythos system card highlights a stark capability jump: Opus 4.6 achieved near-zero success at autonomous exploit development, while Mythos successfully created 181 working JavaScript shell exploits. Specific discoveries include a 27-year-old OpenBSD TCP vulnerability and Linux privilege escalation exploits via race conditions. Willison calls the restricted release 'necessary' and frames it as an 'industry-wide reckoning in the making' for software security.