Anthropic Launches Free AI-Powered Open-Source Security Scanner

Anthropic has introduced OSS Scanner, an opt-in AI-driven vulnerability scanner designed to secure open-source projects. The service offers free, periodic security scans using advanced models like Claude Mythos, with no human review required. Project maintainers can enroll by submitting a pull request with a YAML configuration file detailing repository information and contact details. The scanner aims to identify and report vulnerabilities without imposing a strict disclosure period, though this may change as confidence in the tool grows. Anthropic has already found over 29,000 potential vulnerabilities, reporting more than 6,000 flaws to maintainers, resulting in 584 advisories. This launch coincides with the company’s Critical Infrastructure Defense Program, aimed at bolstering cybersecurity defenses in critical infrastructure and open-source software. Anthropic predicts that AI will increasingly favor defense, making it easier to catch bugs and write secure software.