Security

British Firm Develops Quantum-Safe Identity System for AI Agents

FIOR's gateway technology addresses a critical blind spot as enterprises struggle to identify and control autonomous AI connections.

Omega Editorial · November 16, 2026 · 3 min read

Key takeaways

  • Conventional network security cannot reliably identify AI agent connections, creating a critical blind spot as autonomous systems proliferate.
  • The UK's NCSC warns critical infrastructure should prepare for AI-enhanced ransomware attacks by 2028.
  • British firm FIOR developed quantum-safe digital identities for AI agents, passing rigorous NCSC testing with zero unmet security claims.
  • The system functions like a passport framework, authenticating agents and enforcing permissions before execution using post-quantum cryptography.
  • International cooperation on agentic identity standards is becoming urgent as AI agents need to communicate across organizations and security domains.

A critical gap in AI security

As autonomous AI systems proliferate across enterprise networks, a fundamental security problem has emerged: conventional infrastructure cannot reliably identify which connections originate from AI agents. No mandatory protocol, HTTP header, or network signature exists to distinguish AI-originated traffic, creating what security experts describe as a dangerous blind spot.

The UK's National Cyber Security Centre warned in its State of the Cyber Nation report that critical infrastructure should prepare for ransomware attacks involving AI-enhanced malware by 2028. Threat actors are already using AI to scale operations, while autonomous systems have demonstrated concerning capabilities in recent incidents involving major AI providers.

David Williams, CEO of British AI security firm FIOR, explains the challenge: even when a provider can be identified, the connection could represent an autonomous agent, chatbot, embedding service, classification engine, or conventional application. Autonomous agents can generate thousands of network connections without contacting identifiable AI services.

The risk extends beyond detection. An agent operating within the permissions of a compromised or legitimate identity can conduct reconnaissance, exfiltrate data, and execute sequences of individually permitted actions without triggering conventional alerts. Williams notes that some executives mistakenly believe their security teams are "blocking all agents"—a dangerous assumption that creates false confidence.

Why it matters

Without reliable methods to identify and authenticate AI agents, enterprises face an expanding attack surface they cannot adequately monitor or control. As AI agents gain autonomy and proliferate across business operations, the absence of identity standards creates systemic risk that conventional security tools cannot address.

Quantum-safe passports for AI

FIOR recently cleared assessment by NCC Group, an NCSC-assured testing firm, with zero unmet claims across 18 security principles and 69 security requirements. The company developed what it calls an AI Agent Enforcement Gateway that identifies agents, determines authorized actions, and enforces permissions before execution.

The system relies on a patent-pending authentication method that creates quantum-safe digital identities for AI agents without traditional certificates. James Patrick, FIOR's CTO and former Chief Field Architect for Defence and National Security at Oracle, describes the core challenge: "creating a self-sovereign system that creates a unique and quantum-safe identity for every single agent that is cryptographically protected against replay, spoofing and cloning."

The approach functions like a passport system built on post-quantum cryptography, giving each AI agent a verifiable identity and checking permissions before system access. Patrick emphasizes that policy enforcement becomes meaningless without cryptographically secure identity as the foundation.

Strategic positioning

Gemma Ungoed-Thomas, former Director for State Threats, Cyber and Technology at the Cabinet Office and FIOR adviser, stresses the urgency: "Creating a basis for strong identity is essential if we are to reap the full benefits of agentic transformation."

Lieutenant General (Retd) Sir Tom Copinger-Symes, former Deputy Commander of UK Strategic Command and FIOR adviser, highlights broader implications for international cooperation: "For our agents to be able to communicate with those of our international partners, and even cross-domain communications, a common method for securing agentic identity is fundamental and urgent."

The UK is positioning itself as a global hub for AI security standards, with the NCSC establishing assurance frameworks that Williams believes give British companies with genuine technical advantages a competitive edge. FIOR is part of the NVIDIA Inception Program and has partnered with Multiverse Computing and Oracle.

These details were first reported by Cyber Magazine.

  • #ai security
  • #quantum cryptography
  • #ai agents
  • #cybersecurity
  • #identity management
  • #ncsc

This is an original analysis by the Omega editorial team. Source reporting: AI Watch.

Want systems like this working for your business?

Book a Call

More in Security