Verizon joins Anthropic’s Project Glasswing to run the Claude Mythos Preview model on its network. The AI, designed to spot complex software vulnerabilities at speed, will let Verizon tighten its security posture.

Anthropic set up Project Glasswing for a select group of global security leaders. Verizon, the first telecom to access Mythos, will use it to identify and contain hard‑to‑find flaws while upholding strict safety protocols.

CEO Dan Schulman said the initiative helps protect customers and “share cross‑industry insights that will help secure the global internet fabric.” By testing Mythos, Verizon aims to strengthen its high‑standard network defense.

The partnership marks a rare move for a carrier, positioning Verizon to lead industry best practices in AI‑driven threat detection and to safeguard critical infrastructure.