Anthropic Launches AI-Powered Bug Reporting for Open Source and Critical Infrastructure Security Initiatives
Anthropic has introduced two new cybersecurity initiatives aimed at enhancing vulnerability detection and remediation. The first, OSS Scanner, is a free service that utilizes Anthropic's advanced AI models to periodically scan open-source projects for potential vulnerabilities. Maintainers who opt-in receive model-generated reports, including proof-of-concept exploits and suggested fixes, directly and without human review, to expedite the disclosure process. Anthropic anticipates a true-positive rate exceeding 90% for these reports. The second initiative, the Critical Infrastructure Defense Program (CIDP), focuses on operational technology (OT) security. This program partners Anthropic's frontier Claude models and on-site engineers with 11 founding firms, including Accenture, Booz Allen, and Deloitte, to address the unique challenges of securing critical infrastructure sectors like power, water, manufacturing, and transportation. These firms will leverage Anthropic's AI and threat research to help customer...
