Anthropic has introduced OSS Scanner, a free service designed to identify vulnerabilities in open-source software projects. The tool leverages the company's advanced AI models, specifically Claude Mythos, to automate the detection process.
Participation in the service is based on an opt-in model. Users who subscribe will receive automated reports regarding potential security flaws identified by the AI system without human oversight.
The company warns that the reports are generated entirely by AI, meaning they may contain false positives or invalid findings. Developers are responsible for manually verifying all results provided by the scanner.
The launch follows a trend of using AI to secure software, which previously helped identify critical issues like the Copy Fail vulnerability in Linux distributions. However, this automation increases the volume of reports.
Increased automated reporting has created challenges for maintainers, including Linus Torvalds and Google, who are struggling to manage the influx of AI-generated error reports and security notifications.
