‘AI is changing cyber security fast’: BT becomes first UK firm to join Anthropic Project Glasswing

BT is the first UK firm to join Anthropic’s Project Glasswing
High-risk Claude Mythos Preview model is limited to select partners only
UK infrastructure set for major security boost – BT already blocks 4m attacks daily

BT has become the first UK company to publicly confirm membership of Anthropic’s Project Glasswing, a cybersecurity initiative that grants partners access to the company’s most advanced cybersecurity model ,Claude Mythos Preview.

The announcement was made at the UK Government’s AI Adoption Summit, where BT CEO Allison Kirkby declared the partnership would help the company to defend both its own networks and customer systems against evolving and sophisticated attacks.

“AI only works at scale when it is underpinned by future-ready networks that are secure, resilient, safe,” Kirkby said, committing to “working with Government to support the further development and deployment of sovereign British AI capability, so that the UK can be an AI maker and not just a taker.”

BT joins Anthropic’s Project Glasswing

Anthropic launched Project Glasswing in April 2026 with select partners in response to AI-fuelled attacks.

At the time, Anthropic boasted that Claude Mythos Preview had already found thousands of high-severity vulnerabilities spanning every major OS and browser.

While the frontier model has proven benefits across identifying previously unknown software vulnerabilities (including 16- and a 27-year-old vulnerabilities), generating potential exploit paths and recommending security patches, it remains unreleased to the general public over concerns that it could be misused by attackers.

Many regard it as the current epitome of AI fighting AI, or fighting fire with fire, putting the most capable and powerful model up against the rising volume and sophistication of attacks that AI itself has enabled.

Because BT’s networks underpin large portions of the UK’s infrastructure, it means that getting on-board will have far-reached benefits for consumers and businesses. The company boasted that it now prevents four million cyberattacks across its networks every single day, and that’s before it gets on-board with Project Glasswing.

“By joining Project Glasswing, BT will strengthen its own cyber security capability to protect our networks, our customers and the wider UK,” BT Business CEO Jon James added.

Inside Anthropic’s plans to expand Project Glasswing

A few weeks after the project launched, more than 10,000 high- or critical-severity vulnerabilities have been identified.

On June 2, Anthropic expanded Mythos access to 150 new organizations across more than 15 countries, which reportedly include Australia, Canada, France, Germany, Italy, Switzerland, the Netherlands, Spain, Belgium, Sweden, India, Japan, New Zealand and South Korea (per The Times of India).

Participants now span telecoms, energy, healthcare, government and more. At launch, the Claude maker already declared it was “in ongoing discussions with US government officials” about how Clade Mythos Preview could help its offensive and defensive strategies.

“Governments have an essential role to play in helping maintain that lead, and in both assessing and mitigating the national security risks associated with AI models,” the company added, expressing an appetite to work with local, state and federal representatives.

Looking ahead, Anthropic predicts that developers could gain access to similarly capable models within the next six to 12 months with adequate safeguards in place so that they don’t amplify the threat they’re designed to tackle.

Separately, the Government revealed that BT had agreed to “share data and insights on how [it’s] using AI in the workplace” to help guide broader workplace rollouts