Skip to the content
AI and Cybersecurity Learning and Practice Center Test automation Open the Learning Center

NewsProduct

Anthropic opens a free AI vulnerability scanner to open-source maintainers, with reports sent unreviewed

Anthropic launched OSS Scanner, an opt-in service that gives core maintainers of eligible open-source projects periodic scans by its strongest models, at no cost. Projects enroll by pull request to the anthropics/oss-scanner repository and are chosen case by case, with criteria similar to Google's OSS-Fuzz, for critical impact on infrastructure and user security. Each report has a self-contained reproducer, an explanation and, when available, a candidate patch, and is fully model-generated with no human review, so some will be wrong and maintainers stay responsible for validating them. Anthropic says it has found over 29,000 candidate vulnerabilities and sent nearly 5,000 reports; when penetration testers checked 97 critical and high findings across 48 projects, 85 met its disclosure bar, 11 were real but duplicates and one was a false positive. Human-verified reports still go through its coordinated disclosure process.

Why it matters

Finding bugs is getting cheaper than fixing them. Maintainers who opt in trade a human filter for speed, so they need the capacity to triage a steady stream of reports; everyone else should expect more fixes, and more CVEs, in the open-source code they depend on.

Sources

Read it with the rest of the news in the Learning Center