Milan's Beelzebub raises €3M to counter AI-driven cyberattacks

The startup addresses attacks in which adversaries use AI to automate vulnerability discovery, generate custom malware and launch large-scale operations at machine speed. According to Tech.eu, Beelzebub operates on an assumed-breach model, detecting malicious activity from attackers already inside a network rather than relying only on perimeter defences. Its platform combines attack simulation, deception technology and automated threat analysis.

Tech.eu reported that the platform has three integrated components: Arcangelo, which simulates targeted attacks to test defences; Beelzebub Managed, which deploys AI-powered decoy infrastructure to attract and detect intruders; and Caronte, an AI malware analysis component.

Beelzebub began as a research project by Mario Candela in Milan in 2021 and was incorporated in July 2025. Candela, the company's CEO and founder, told EU-Startups that AI-powered attackers have become too fast and numerous for humans to counter alone, and that the round would accelerate efforts to bring modern cybersecurity to companies that cannot afford to compromise.

Beelzebub also noted the growing regulatory pressure across Europe from frameworks such as the NIS2 Directive and the Cyber Resilience Act.