Nozomi Networks uses Claude through Project Glasswing to find and fix vulnerabilities in its own platform, according to CTO Moreno Carullo. The review is now part of the standard security process, he wrote in the company blog on October 8, 2026.
His account accompanies Nozomi's participation in Anthropic's Critical Infrastructure Defense Program. It identifies a specific security practice beyond the partner quote: AI is intended to support platform reviews and analysis of research findings.
Carullo stresses that people retain responsibility for decisions and actions in operational environments. For industrial equipment with long lifetimes and rare maintenance windows, AI can assist specialists but should not replace their judgment.
The post gives no precise implementation date, published finding counts or independent effectiveness measurements. October 8 dates the public statement. Confirmed membership does not establish that every described safeguard is a new feature of this program.