Italy's Beelzebub secures €3M to counter AI-driven cyberattacks

Beelzebub works on an assumed-breach principle rather than depending only on perimeter defences, aiming to spot malicious behaviour from intruders that are already within a network. Its platform brings together attack simulation, deception technology and automated threat analysis to detect, isolate and respond to threats.

The system is made up of three connected components. Arcangelo runs simulated targeted attacks to test an organisation's defences on an ongoing basis, while Beelzebub Managed sets up AI-powered decoy infrastructure to lure and identify attackers inside a network. A third component, Caronte, is described as the platform's AI malware element.

Tech.eu reported that demand for stronger defences is rising as European regulatory frameworks such as the NIS2 Directive and the Cyber Resilience Act increase security obligations.

Anthropic Says Russia- and China-Linked Actors Used Claude in Espionage and Weapons Work

The AI company reports it detected and disrupted state-linked groups using its Claude models for hacking campaigns and military development projects.