Anthropic Expands Access to Latest AI Models for Cyber Firms

Anthropic is broadening its Cyber Verification Program to grant authorized security professionals access to Claude models with reduced safety restrictions. This initiative allows vetted organizations to utilize the technology for vulnerability research, penetration testing, and malware analysis. By collaborating with the US government and security partners, the company aims to help identify systemic software weaknesses at scale. Previous testing through this program successfully uncovered over 129,000 security flaws, demonstrating the potential for these models to assist in large-scale infrastructure defense.