Anthropic Launches Free AI Security Scanning for Open-Source Projects
The newly announced OSS Scanner service provides participating projects with periodic, model-generated security reports using powerful models like Claude Mythos.
Anthropic has announced the launch of a new service called OSS Scanner, aimed at helping open-source projects identify security vulnerabilities.
Details of the OSS Scanner Service
Open-source projects that join the system will be able to benefit from comprehensive, periodic security scans performed by Anthropic's most powerful models at no cost.
Model-Generated Reporting Process
The outputs of this optional security scanner are generated entirely by the model without human review or triage, thereby enabling much faster and more frequent scanning.
Anthropic specifically notes that these generated reports may potentially be flawed or invalid.
Claude Mythos and Security Advantages
The reports are generated using advanced models, including Claude Mythos, in order to provide defensive advantages to open-source projects.