Anthropic released Claude Code Security as a limited research preview for Enterprise and Team customers, with expedited free access for open-source maintainers. The tool reasons about data flow across a codebase, re-examines each finding in a multi-stage verification pass, assigns severity and confidence ratings, and proposes patches that are applied only with human approval.
Why it matters
It packages frontier-model vulnerability finding for defenders with explicit human approval gates, amid concerns about the same capability aiding attackers.
Key facts
As stated in the sources, with where to find them.
- Released as a limited research preview to Enterprise and Team customers, with expedited access for open-source maintainers.Opening and 'Getting started'
- Every finding goes through multi-stage verification where Claude tries to prove or disprove it; nothing is applied without human approval.Section 'How Claude Code Security works'
- Anthropic restates that Opus 4.6 found over 500 vulnerabilities in production open-source codebases.Section 'Using Claude for cybersecurity'
Findings that cite this record
No tracked finding cites this record yet.
Key questions this bears on
- Is AI shifting the balance between finding and fixing vulnerabilities?Discovery is ahead. Verification, triage, coordinated disclosure, safe repair and getting fixes to users remain bottlenecks between AI findings and protection.
- Can AI agents defend and oversee systems on their own?Not reliably. Some designs improve benchmark triage, but autonomous defense lacks operational validation and production oversight of agents has documented gaps.
Sources
Related records
Feb 5, 2026
May 22, 2026
Apr 7, 2026
Mar 6, 2026
Oct 3, 2025
Oct 8, 2026