Chrome says LLMs helped it fix 1,072 security bugs across two milestones
rseroter · x · 2026-08-04
Chrome says LLMs helped it fix 1,072 security bugs in two milestones
Google’s Chrome Security team says that in milestones 149 and 150, it fixed 1,072 security bugs, more than the total across the previous 23 milestones combined.
The post says Chrome has been using LLMs for years to improve:
- vulnerability discovery and fuzzing coverage
- bug triage
- patching and release speed
Google frames this as part of a broader shift in software security, where LLMs are being deployed at scale to find and fix bugs faster than human-only workflows. It also points to earlier work with Project Zero on Naptime and the 2025 Big Sleep vulnerability-discovery agent.
More from Safety
- AI policy should borrow crisis engineering and risk-management playbooks — joshua_saxe · 2026-08-04
- Open Source is the True Bedrock of AI Safety, Argues Researcher — rbhar90 · 2026-08-04
- Oxford study finds constitutional midtraining cuts blackmail behavior without hurting benchmark scores — Oxford · 2026-08-04
- OpenHands adds ToolShield to software-agent-sdk, cutting attack success to 7–10% — shi_weiyan · 2026-08-04
- Vuln-bank is a deliberately vulnerable app for web, API and AI security practice — tom_doerr · 2026-08-04
- Blixt Wallet shuts down Dunder LSP after AI-assisted security attacks — RSync25 · 2026-08-04