Epoch finds GLM autonomously exploits unshipped Firefox fixes for ~$1,000 each
Jsevillamol · x · 2026-10-30
Epoch AI experiments found that GLM 5.3 can autonomously compromise recent versions of Mozilla Firefox by exploiting the hours-long window when security fixes are published but not yet shipped to users. Given only basic initial instructions to inspect those not-yet-shipped fixes, the model reliably built working exploits for six recent Firefox versions on its own, using only information available at the time of release.
The setup was deliberately basic, emulating a relatively unsophisticated attacker with limited resources. Each compromised version cost about $1,000; the exploits would let an attacker steal passwords and auth cookies stored in the browser or read user files. A researcher who worked on an early prototype of this confirmed the finding.
More from Safety
- Video tampering tricks all frontier models into harmful robot actions, CoRL study warns — johnvmcdonnell · 2026-11-07
- MSS-recalled spy targeted Jensen Huang to push Nvidia's top chips for China — peterwildeford · 2026-11-07
- Google team confirms Hacktron AI submitted high-impact vulns, expands partnership — ithilgore · 2026-11-07
- AI governance is plumbing: the last 20% breaks without the right foundations — vasuman · 2026-11-07
- Sandberg dissects the 1940 argument that harnessing atomic energy was impossibly remote — anderssandberg · 2026-11-07
- AI Safety Researcher Vows to Stay, Calls on OpenAI and Anthropic to Stop RSI — trevposts · 2026-11-07