Artificial intelligence company Anthropic has launched a new service called OSS Scanner, providing free automated security scans to open-source software projects. By opting into the program, developers can utilize Anthropic's most advanced AI models—including Claude Mythos—to check their code repositories for security vulnerabilities.
The service is designed to deliver rapid, periodic security reports without human intervention or triage. However, Anthropic noted that because the scanning process relies entirely on AI models without human review, some generated vulnerability reports might be incorrect or invalid. Despite the potential for false reports, the company aims to give open-source projects a stronger defensive advantage against security flaws.
AI tools have previously helped spot major security vulnerabilities, such as the "Copy Fail" bug that affected nearly every Linux distribution earlier this year. However, the sudden rise of AI-generated reports has also strained developers. Major tech organizations including Google, as well as Linux creator Linus Torvalds, have recently struggled to manage the overwhelming volume of automated bug reports submitted to open-source projects.