Seventeen years of enterprise security,applied to AI systems.

Nicholas has spent seventeen years inside the security organisations of European enterprises - designing perimeter architectures, implementing zero-trust programmes, and hardening the systems that handle money, energy, and regulated data at scale.

Rogue AI was founded to apply that same discipline to a new class of system. Large language models, agentic frameworks, and retrieval pipelines are now being deployed into the same regulated environments that took two decades to secure - often by teams who have never operated production infrastructure under regulatory scrutiny. The practice exists to close that gap.

“The interesting work in 2026 is not getting AI to function. It is getting AI to function safely, reliably, and compliantly inside the same environments we’ve been defending for years. That is an architecture problem before it is a model problem.”