Beelzebub, an Italy-based cybersecurity startup, has raised a €3 million seed round led exclusively by United Ventures. Combined with a prior €300,000 pre-seed from strategic investors and advisors, the company's total funding now stands at €3.3 million.
The Problem: AI Is Arming Attackers
The premise behind Beelzebub is straightforward but urgent. AI tools have dramatically lowered the cost and skill floor for launching sophisticated cyberattacks — enabling automated vulnerability discovery, custom malware generation, and large-scale intrusions that operate at machine speed.
At the same time, European regulatory pressure is intensifying. Frameworks like the NIS2 Directive and the Cyber Resilience Act are raising the compliance bar for organisations across the continent, creating both obligation and incentive to modernise their defences.
Assumed Breach, Not Just Perimeter Defense
Beelzebub's core architecture departs from traditional perimeter-focused security. Instead, it operates on an assumed-breach model — treating network infiltration as an inevitability and focusing on detecting and containing threats already inside.
The platform is built around three integrated modules:
- Arcangelo — continuously simulates targeted attacks to stress-test an organisation's defences
- Beelzebub Managed — deploys AI-powered decoy infrastructure (deception technology) to attract and expose attackers operating within a network
- Caronte — an AI malware analyst that automatically reverse-engineers malicious software and generates incident reports, available both cloud-hosted and fully on-premises
The on-premises option is a notable differentiator for enterprises in regulated industries with strict data residency requirements. The platform is also designed to satisfy NIS2 compliance requirements directly.
What the Funding Buys
Founder and CEO Mario Candela framed the raise around a fundamental shift in the threat landscape:
"The rise of AI-powered attackers has fundamentally changed cybersecurity, requiring organisations to complement human expertise with AI-based systems capable of responding at machine speed. Beelzebub's product adapts to new types of malware and is always updated to match the current state of the most sophisticated attacks. This seed round will supercharge our efforts to bring modern cybersecurity to the companies who cannot afford to compromise."
The €3M will be deployed across four priorities:
- Expanding the research and engineering team
- Opening commercial offices in Rome and San Francisco before end of 2025
- Accelerating customer acquisition across Europe, with emphasis on NIS2-regulated organisations
- Developing new technologies to protect AI agents directly — an emerging attack surface as agentic AI systems proliferate in enterprise environments
Why This Matters for Enterprise Buyers
The market for AI-native security tooling is crowding fast. Competitors ranging from CrowdStrike and SentinelOne to a wave of European startups are all racing to embed AI deeper into detection and response workflows. What distinguishes Beelzebub's positioning is the explicit focus on the assumed-breach model combined with deception technology — a combination that's well-suited to catching sophisticated attackers who evade signature-based detection.
The San Francisco office signals an intent to compete beyond Europe, though the near-term commercial thesis is clearly anchored in the regulatory tailwinds of NIS2. For enterprise security buyers under compliance pressure, a platform that bundles attack simulation, internal threat detection, and automated malware analysis into a single SaaS website-style deployment model may simplify procurement considerably.
With AI agents becoming a new attack surface — as enterprises deploy autonomous systems with broad access permissions — Beelzebub's roadmap investment in agent-specific protections positions it ahead of a problem that most security vendors haven't fully addressed yet.
