Netskope Joins Anthropic’s Project Glasswing

Netskope said it joined Anthropic’s Project Glasswing, using Anthropic’s Claude Mythos Preview (Mythos Preview). The company tied the effort to finding vulnerabilities in code, with findings intended to be shared within the Glasswing coalition and the cybersecurity community.

The release described a shift in AI usage, stating that an average enterprise saw the number of AI applications in use grow fivefold in the past year and tripled its AI user base. It also cited more than 220 AI data policy violations per month and said more than 94% of organizations reported gaps in AI activity visibility.

Netskope stated that it held access to Claude Mythos Preview and that Project Glasswing used Mythos Preview to find vulnerabilities in code. The company also described its role as “the global inline inspection point for AI and all enterprise traffic,” processing “trillions of transactions.”

Netskope said it planned to continue sharing findings with the broader Glasswing coalition and the cybersecurity community. Sanjay Beri, co-founder and CEO of Netskope, said, “Netskope is the global inline inspection point for AI and all enterprise traffic for our customers, processing trillions of transactions across the world's largest organizations,” and added, “The architectures and ecosystems that will carry enterprises forward are the ones built for real-time governance, and security that understands the language of the AI and cloud world. With that in mind, we are proud to be collaborating with Anthropic on Glasswing and other important initiatives.”

The release also stated that Netskope would continue collaborating with leading frontier AI models on work intended to secure and defend organizations at AI speed.