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

Playing the video loads YouTube's player, which sets cookies. Nothing is loaded from YouTube until you press play.

Organised crime groups and state-linked actors are using artificial intelligence to accelerate operations and create malicious tools. Following a wave of these attacks in 2026, security teams need improved incident response procedures and communication protocols. Researchers Dr Louise Marie Hurel and Daniel Cuthbert examine these defensive and offensive uses of AI in a commentary for RUSI. They also highlight the need to understand how AI laboratories work with evaluator companies such as METR, which test, grade and audit the models.

Summary written automatically from the publisher's own text. The original video is the record.

Watch on YouTube

Related coverage

More on automated decisions