Anthropic's Claude identifies 493 security issues in open-source projects

Anthropic's Frontier Red Team deployed Claude AI models across 25 major open-source projects, analyzing over five million lines of code to identify 493 attack surface issues. The research uncovered critical configuration flaws, including a 21-year-old authentication bypass and unauthenticated exposure in OpenSSH 10.4, prompting maintainer patches. Alongside these findings, Anthropic updated Claude Security to introduce commit-driven continuous repository monitoring for enterprise users.

Frontier Red Team open-source research

• ▪The Frontier Red Team's research on attack surface reduction was written up by Anthropic security engineer Chris Rohlf and published on October 19, 2026.
• ▪Anthropic's Frontier Red Team conducted research using Claude models across 25 widely used open-source software projects, identifying 493 issues related to attack surface reduction.
• ▪Claude scanned over five million lines of C, C++, Java, PHP, and Python code across 25 open-source projects, inventorying over 17,000 configuration options, with 3,842 touching security controls.
• ▪Anthropic introduced an opt-in OSS Scanner for open-source projects, utilizing Claude models to address vulnerabilities across 25 projects and surface 493 security issues.

Vulnerabilities found in open-source projects

• ▪In four unrelated open-source projects, Anthropic found a configuration flaw where host-based deny rules fail open when reverse DNS lookups are disabled, allowing blocked clients to access the server.
• ▪Anthropic identified a 21-year-old configuration conflict in an open-source project where an option meant to delay client certificate checks silently disabled the requirement, allowing unauthenticated connections.
• ▪In 11 gated open-source projects analyzed by Anthropic, Claude identified code reachable prior to user login, including 93 unnecessary pre-authentication surfaces and 37 places running code or saving state before authentication.
• ▪Claude identified 162 security-sensitive configuration option combinations in open-source projects that software accepted without error or warning, with 94 combinations silently disabling a security control.

OpenSSH vulnerability findings and fixes

• ▪Following Anthropic's report and provided patches regarding post-quantum signature code exposure in OpenSSH, project maintainers released fixes across OpenSSH releases 10.5 and 10.6.
• ▪In the OpenSSH 10.4 release, Claude traced the SSH daemon and identified approximately 4,500 lines of post-quantum signature code reachable by unauthenticated clients by default, representing about 16% of pre-authentication code.

Claude Security features and public beta

• ▪Claude Security continuous monitoring is available in public beta for Claude Enterprise customers, with eligible organizations offered up to $25,000 in monthly usage credits for two months through December 31, 2026.
• ▪Since entering public beta in April 2026, Claude Security has been utilized by thousands of organizations to identify over one million vulnerabilities.
• ▪Claude Security supports connected projects with up to eight repositories across platforms like GitHub and GitLab, allowing findings to integrate with external tools via API and Model Context Protocol (MCP) standards.
• ▪Anthropic announced an update to Claude Security on October 21, 2026, introducing continuous monitoring that automatically rescans connected code repositories based on new commit triggers rather than calendar schedules.