Beelzebub Raises €3M for AI-Native Cyber Defense

Most cybersecurity products are built to stop an attacker at the network perimeter.

Beelzebub starts from a less comfortable assumption: the attacker may already be inside.

The Milan-based cybersecurity startup has raised €3 million in Seed funding, led exclusively by United Ventures. The round follows an earlier €300,000 pre-seed investment and brings the company’s total funding to €3.3 million.

Founded by Mario Candela, Beelzebub is building an AI-native cybersecurity platform for companies facing faster and increasingly automated attacks. The company argues that attackers are using AI to identify vulnerabilities, generate custom malware and operate at a speed that human security teams cannot consistently match.

Its platform is built around three connected products. Arcangelo runs targeted attack simulations to test whether an organization’s security controls can withstand realistic threats. Beelzebub Managed deploys AI-powered decoys across cloud, infrastructure and application environments. These systems are designed to appear valuable to an intruder while remaining isolated from production assets.

When an attacker interacts with one of those decoys, the platform collects information about the intrusion and tracks lateral movement inside the network. Beelzebub states that this approach can reduce alert noise because legitimate users should have little reason to access a decoy environment. Its platform supports cloud systems, Kubernetes, APIs, IoT devices and AI-agent environments.

The third product, Caronte, works as an AI malware analyst. It examines malicious software, reverse-engineers its behavior and produces incident reports. The tool can operate in the cloud or on-premises, which is important for organizations that cannot send sensitive files or security data to an external service.

Beelzebub is entering the market as European companies face stricter security and reporting requirements under the NIS2 Directive and the Cyber Resilience Act. The company is positioning its SaaS and on-premises products for organizations that need stronger post-breach visibility and clearer audit trails.

The funding will be used to expand Beelzebub’s research team, open commercial offices in Rome and San Francisco, and increase customer adoption across Europe. The company is placing particular emphasis on organizations affected by NIS2 and also plans to develop security technology designed specifically to protect AI agents.

The approach is practical, but it will need to prove itself inside large enterprise environments. Deception systems must remain convincing enough to attract attackers without becoming difficult to manage. Automated malware analysis must also produce findings that security teams can trust during a real incident.

Beelzebub is not promising to eliminate every breach. Its model is based on detecting what happens next and responding before an attacker can move further.

Source : Tech.eu