Verizon claimed to be the first telecom operator to join Project Glasswing, the elite AI-driven bug hunting taskforce put together by AI vendor Anthropic.

Verizon joins the likes of Cisco, Nvidia, Broadcom, and Palo Alto Networks in Project Glasswing, with the telecom operator claiming the move would help it identify and remediate complex vulnerabilities while maintaining its "high standards of network protection."

Project Glasswing was established in April and centers around Claude Mythos Preview, Anthropic's new large language model (LLM). Anthropic claimed Mythos Preview to be a general-purpose LLM with a particular knack for cybersecurity tasks, having found “thousands” of high-severity vulnerabilities across every major operating system and web browser without human steering.

With no current plans to make Mythos publicly available, its use is currently limited to Glasswing partners as well as a group of more than 40 additional and unnamed organizations.

Verizon CEO Dan Schulman noted Verizon as the "only" telecommunications company utilizing the model, but this was disputed by AT&T following the announcement.

"We are participating in Project Glasswing, an initiative to strengthen critical software using Claude Mythos Preview," an AT&T spokesperson told SDxCentral. "We have long been committed to innovation that helps keep our network available and secure for the customers who rely on us every day. This includes cutting-edge AI programs, sophisticated network tools, and collaboration with industry partners."

Commenting on its partnership, Schulman said: “Our customers rely on the security of our network every day. As part of Project Glasswing, we are able to test and improve our cybersecurity efforts with new insights to maintain our network’s security. Over the past several months, our information security team has been rigorously testing this critical new technology to determine its benefits to our network. We are uniquely positioned to share cross-industry insights that will help secure the global internet fabric and support our mission to deliver a safe and reliable experience for every customer.”

Verizon added it "adheres to strict usage policies ensuring that all emerging technologies are evaluated safely and responsibly," with every phase of the Mythos Preview testing process "governed by rigorous safety standards."

Project Glasswing's elite nature aims to keep its advanced Anthropic AI out of bad actors' hands. As recently explored by this title, AI is being explored both by enterprises and cyber attackers, with Nokia warning that state-sponsored threats are adopting AI in their arsenal, leading to a need for AI-driven response speeds in distributed denial of service (DDoS) defense services.

Samantha Kight, head of industry security at telecom industry group GSMA, recently told SDxCentral that AI has already greatly helped telecom-based security operation centers.

"It helps them be able to link different indicators of compromise to one another," Kight said. "It helps them generate reports to present the best advice against a threat. So there's been so many advantages to it, but as a result, they're spending a lot of time also having to defend against AI attacks.”

Story updated to include comment from AT&T.