The Double-Edged Sword: AI in Cybersecurity Offense and Defence

In this video commentary, Dr Louise Marie Hurel, Senior Research Fellow in Cyber and Tech and Daniel Cuthbert, RUSI Associate Fellow, discuss how organised crime groups and state-linked actors are leveraging AI to speed up operations and build malicious tools, as well as the need for better incident response procedures and communication protocols following a wave of such attacks in 2026.

How can we better understand the relationship between AI labs and evaluator companies such as METR, who test, grade and audit them?

RUSI's Cyber and Tech 'Secure Access to Frontier AI' project taskforce explores the cyber security, information security and model-level risks associated with third-party access to frontier AI models.

The team's paper, released in May 2026, delivers a critical framework for securely enabling third-party evaluation of frontier AI models, directly addressing the urgent need for robust safety protocols in the defence and security sector as AI capabilities rapidly advance.