Anthropic Launches Free AI Scanner to Find Open-Source Security Bugs

Anthropic has introduced OSS Scanner, a free, opt-in service that uses its AI models to identify potential vulnerabilities in eligible open-source repositories.

Participating projects receive periodic scans at no cost.

Reports and Proposed Fixes

The service emails maintainers bundles of findings. Its scanning process includes AI agents that check potential bugs, analyse their causes and propose patches.

Unlike Anthropic’s standard coordinated vulnerability disclosure process, these reports arrive without prior human review. Maintainers must therefore assess findings for errors or false positives.

Who Can Apply?

Anthropic prioritises established projects with significant importance to infrastructure or user security. Applications are evaluated individually, and the company verifies that applicants are core maintainers.

How Maintainers Join

Maintainers can apply through Anthropic’s OSS Scanner GitHub repository by submitting the required project configuration and build environment.

The service offers a faster route to potential findings, while leaving maintainers responsible for validating reports and deciding which fixes to adopt.