Instead of building a stronger perimeter, Beelzebub assumes an attacker is already inside the network and deploys realistic digital decoys—LLM-powered traps that look like real servers, databases, APIs, cloud systems, and IoT environments—to lure attackers into revealing themselves .

How the Platform Works: Three Core Capabilities

Beelzebub’s platform runs as a single automated loop across three integrated products :

• Arcangelo – The offensive simulation layer continuously simulates attacks from a frontier-model adversary against the live perimeter, flagging newly disclosed vulnerabilities in real time .
• Beelzebub Cloud – Deploys LLM-powered decoy infrastructure (fake servers, databases, APIs) across the network. When an attacker touches a trap, the platform detects and blocks the attack, isolates affected machines, and triggers response actions .
• Caronte – Reverse-engineers captured malware and automatically produces incident reports, closing the loop between detection and remediation .

The platform is offered as SaaS or on-premise for the most sensitive environments, including LLM inference on Beelzebub-supplied hardware, and is described as NIS2-ready .

Funding Details and Investor Backing

Beelzebub raised a €3 million seed round led exclusively by United Ventures, an Italian deep-tech VC. This follows a €300,000 pre-seed round from strategic investors and advisors, bringing the company’s total funding to €3.3 million .

United Ventures’ investment thesis hinges on what they call “the symmetry is gone” — the idea that AI-powered attackers can strike faster and at a scale that perimeter defenses cannot match . Beelzebub’s assumed-breach model is designed for that new reality.

What the New Funding Will Be Used For

Beelzebub plans to deploy the capital across four areas :

1. Expand the research team – Particularly in AI and cybersecurity R&D .
2. Open commercial offices in Rome and San Francisco – By the end of the year .
3. Accelerate customer acquisition across Europe – Especially among organizations subject to the NIS2 Directive, which mandates stricter cybersecurity requirements for critical infrastructure operators .
4. Develop technologies to protect AI agents – Defending AI-driven tools from AI-powered attacks .

Target Market: Mid-to-Large Enterprises and Critical Infrastructure

Beelzebub targets mid-to-large enterprises and critical infrastructure operators facing next-generation automated and AI-powered cyber threats . The company’s platform is designed for organizations that can no longer rely on outdated perimeter defenses to keep attackers out.

Why the Assumed-Breach Model Matters

The assumed-breach model is not new in cybersecurity, but Beelzebub’s application of it is distinct. By combining continuous attack simulation, LLM-powered deception, and automated malware analysis into a single automated loop, the platform reduces the time between intrusion and detection — a window that attackers increasingly exploit to cause damage . It essentially turns the network into a trap-filled environment where any attacker movement triggers a response.

Conclusion

Beelzebub represents a growing trend in cybersecurity: moving away from preventing every possible breach toward assuming that some intrusions will succeed and focusing on early detection and automated response. With €3.3 million in total funding, a clear go-to-market plan targeting NIS2-regulated organizations, and offices planned in both Europe and the U.S., the startup is positioning itself as a key player in the next generation of AI-driven cybersecurity defense.

Instead of building a stronger perimeter, Beelzebub assumes an attacker is already inside the network and deploys realistic digital decoys—LLM-powered traps that look like real servers, databases, APIs, cloud systems, and IoT environments—to lure attackers into revealing themselves .

How the Platform Works: Three Core Capabilities

Beelzebub’s platform runs as a single automated loop across three integrated products :

• Arcangelo – The offensive simulation layer continuously simulates attacks from a frontier-model adversary against the live perimeter, flagging newly disclosed vulnerabilities in real time .
• Beelzebub Cloud – Deploys LLM-powered decoy infrastructure (fake servers, databases, APIs) across the network. When an attacker touches a trap, the platform detects and blocks the attack, isolates affected machines, and triggers response actions .
• Caronte – Reverse-engineers captured malware and automatically produces incident reports, closing the loop between detection and remediation .

The platform is offered as SaaS or on-premise for the most sensitive environments, including LLM inference on Beelzebub-supplied hardware, and is described as NIS2-ready .

Funding Details and Investor Backing

Beelzebub raised a €3 million seed round led exclusively by United Ventures, an Italian deep-tech VC. This follows a €300,000 pre-seed round from strategic investors and advisors, bringing the company’s total funding to €3.3 million .

United Ventures’ investment thesis hinges on what they call “the symmetry is gone” — the idea that AI-powered attackers can strike faster and at a scale that perimeter defenses cannot match . Beelzebub’s assumed-breach model is designed for that new reality.

What the New Funding Will Be Used For

Beelzebub plans to deploy the capital across four areas :

1. Expand the research team – Particularly in AI and cybersecurity R&D .
2. Open commercial offices in Rome and San Francisco – By the end of the year .
3. Accelerate customer acquisition across Europe – Especially among organizations subject to the NIS2 Directive, which mandates stricter cybersecurity requirements for critical infrastructure operators .
4. Develop technologies to protect AI agents – Defending AI-driven tools from AI-powered attacks .

Target Market: Mid-to-Large Enterprises and Critical Infrastructure

Beelzebub targets mid-to-large enterprises and critical infrastructure operators facing next-generation automated and AI-powered cyber threats . The company’s platform is designed for organizations that can no longer rely on outdated perimeter defenses to keep attackers out.

Why the Assumed-Breach Model Matters

The assumed-breach model is not new in cybersecurity, but Beelzebub’s application of it is distinct. By combining continuous attack simulation, LLM-powered deception, and automated malware analysis into a single automated loop, the platform reduces the time between intrusion and detection — a window that attackers increasingly exploit to cause damage . It essentially turns the network into a trap-filled environment where any attacker movement triggers a response.

Conclusion

Beelzebub represents a growing trend in cybersecurity: moving away from preventing every possible breach toward assuming that some intrusions will succeed and focusing on early detection and automated response. With €3.3 million in total funding, a clear go-to-market plan targeting NIS2-regulated organizations, and offices planned in both Europe and the U.S., the startup is positioning itself as a key player in the next generation of AI-driven cybersecurity defense.