Before Anthropic's Mythos there was SATAN: what the 1995 scanner panic does and doesn't teach
maier_ak · x · 2026-09-25
Drawing on the 1995 SATAN scanner panic: Dan Farmer and Wietse Venema's tool only matched networks against a static list of known flaws — it couldn't discover new bugs or write exploits. Despite 10,000 downloads in weeks and press warnings of "rocket launchers" for the internet, the predicted collapse never happened. Anthropic's Mythos is categorically different: it reads arbitrary code, finds unknown vulnerabilities and can generate exploits — and access is now controlled by a private firm.
Related event: What the 1995 SATAN Panic Teaches Us About AI Vulnerability Tools(3 posts)→
More from Safety
- AI firms' fair-use defense shaky as their own people call training 'astonishing theft' — rvp · 2026-09-25
- The 'No Magic OOD' Principle: Flagging a Common Flaw in Alignment Plans — nabla_theta · 2026-09-25
- The real test of AI governance: zero egress and independent audit trails on hardware you control — Ghost_Pilot_MD · 2026-09-25
- Content governance is a control problem: without an offline tamper-evident audit ledger, promises stay unverified — Ghost_Pilot_MD · 2026-09-25
- Republican Sen. Todd Young presses White House for formal AI security talks — GaryMarcus · 2026-09-25
- Two distinct AI agent incidents conflated: Transluce AIHW case vs Australia Medicare hack — GaryMarcus · 2026-09-25