BT taps into Project Glasswing in UK first

BT Group has announced it is the first company in the United Kingdom to join Project Glasswing, an initiative led by artificial intelligence firm Anthropic. The collaboration is designed to bolster the telecommunications provider's cybersecurity infrastructure by integrating advanced large language model capabilities into its existing defensive framework. By participating in this project, the operator aims to enhance its ability to identify, analyse, and mitigate sophisticated digital risks that are increasingly driven by automated technologies.

The integration focuses on leveraging Anthropic’s constitutional AI principles to create a more resilient network environment for both consumer and enterprise clients. This move comes as the telecoms industry faces a growing volume of complex cyberattacks that traditional security measures often struggle to detect in real-time. BT expects that the partnership will provide its security teams with more powerful diagnostic tools to oversee vast amounts of network data and flag irregularities with greater precision.

Through Project Glasswing, the operator will gain early access to specific research and software tools developed to counter AI-generated threats, such as automated phishing and malicious code generation. The initiative emphasises the importance of ethical AI deployment, ensuring that the defensive models remain transparent and controllable. This alignment with Anthropic allows BT to test emerging security applications in a controlled environment before deploying them across its broader national infrastructure.

Company representatives stated that the adoption of these tools is a critical component of its wider strategy to modernise its operational efficiency through automation. As the landscape of cyber warfare evolves, the firm believes that staying ahead of bad actors requires the same level of technological sophistication used in the attacks themselves. The project provides a collaborative space where BT can share insights on threat patterns specific to the UK’s telecommunications sector.

The partnership also reflects a broader trend among major service providers to seek out specialised AI firms to address niche technical challenges within their legacy systems. By embedding these intelligent layers into its core operations, the provider seeks to reduce the manual workload on its security operations centres. This shift is expected to decrease response times for critical incidents and provide a more stable service for millions of mobile and broadband subscribers across the country.

Looking forward, BT plans to continue the evaluation of Project Glasswing’s effectiveness as it prepares for further rollouts of AI-assisted security features. The operator intends to use the findings from this initial phase to inform its long-term cybersecurity roadmap and potentially expand its use of generative models in other areas of customer service and network management. Success in this UK-first implementation could set a new standard for how national carriers protect critical national infrastructure against the next generation of digital vulnerabilities.