Claude AI reduces code attack surface in open-source projects

Anthropic's research highlights the capability of frontier AI models, particularly Claude, in automating attack surface analysis and reduction to enhance software security.

Researchers applied Claude to 25 large open-source projects, spanning over five million lines of code in languages including C, C++, Java, PHP, and Python.

The analysis focused on three methods: default configuration, configuration interaction, and early reachability.

OpenSSH case study Claude achieved a notable reduction in the attack surface of OpenSSH, a heavily audited C codebase.

This skill performs three passes on a codebase to identify issues related to default configuration, pre-authentication code, and conflicting security configurations.