Researchers Propose CVE Rate as Safety Threshold for Code Models

Researchers including an Anthropic researcher and a Khosla Ventures partner propose tracking the CVE rate of AI-generated code to set minimum safety thresholds for models used in foundational software, phasing out more vulnerability-prone models.

2026-11-07 ~ 2026-11-07 · 3 related posts

• Could CVE rates set a security bar for code-generating models? — _arohan_ · 2026-11-07
• Proposal: use CVE rates of AI-generated code to set a security bar for models — _arohan_ · 2026-11-07
• A Sweeting proposes CVE-rate security bar for models used in foundational software — _arohan_ · 2026-11-07