Proposal: use CVE rates of AI-generated code to set a security bar for models
_arohan_ · x · 2026-11-07
Arohan proposes tracking which models produced code that later had CVEs, and using those vulnerability rates as a security benchmark to set a minimum bar for models used in foundational software — ruling out older, vulnerability-prone models. Otherwise, he argues, the natural pressure is to optimize purely for cheaper tokens.
Related event: Researchers Propose CVE Rate as Safety Threshold for Code Models(3 posts)→
More from Safety
- Detection is unsolvable yet tools can be highly accurate, researcher argues — CamachoCollados · 2026-11-07
- User shares approval email for Anthropic's Cyber Verification Program Defense Access — AIFlow_ML · 2026-11-07
- Only 1 of 7 browsing agents caught by Cloudflare: the agent identity and authorization gap — Pale_Painting_6271 · 2026-11-07
- Email line-wrap hyphen bug in Apple Mail can spoof sender domains for phishing — kingslyj · 2026-11-07
- Inter-AI Morality: New Paper Explores How AI Systems Treat One Another — dadadavid · 2026-11-07
- Researcher blasts Sanders AI bill: it "condemns millions to die with curable diseases" — basedjensen · 2026-11-07