Anthropic has announced an expansion of its Cyber Verification Program, which now grants a broader range of security professionals access to Claude models with fewer safety restrictions. This initiative is designed to facilitate penetration testing, malware analysis, and vulnerability research.
The decision comes after partners in Anthropic’s previous program identified over 129,000 confirmed vulnerabilities from April to July 2026, including more than 33,000 categorized as high-severity or critical. By providing enhanced access to its AI models, Anthropic aims to strengthen the security landscape and assist professionals in identifying and mitigating potential threats.
Compiled automatically by the Tech AI Newsdesk from public AI-news sources and summarised in our own words.
