FIOR names Gemma Ungoed-Thomas as sovereign AI adviser

FIOR has appointed former Cabinet Office Director for State Threats, Cyber and Technology Gemma Ungoed-Thomas as Adviser on Sovereign AI, adding a senior former UK national security and cyber official to its advisory ranks.

Ungoed-Thomas will advise on sovereign AI strategy, market positioning, national security issues and the public sector implications of emerging technology. Her remit also covers the strategic, security and resilience needs of organisations developing and operating sovereign AI infrastructure.

She spent more than two decades in UK government and most recently served for more than four years in the Cabinet Office role overseeing state threats, cyber and technology. During that time, she helped shape the UK's GBP £2 billion National Cyber Strategy and worked on the country's first Counter State Threats Strategy.

Her responsibilities also included oversight of the Integrated Security Fund, a portfolio of about GBP £160 million supporting cyber and technology work aimed at countering hostile state activity. She also co-chaired the National Cyber Advisory Board.

Earlier, she worked at the National Cyber Security Centre and GCHQ, contributing to the development of the UK's cyber incident response and cross-government cyber defence work. She also served as Private Secretary to the Prime Minister for Home Affairs and National Security.

Alongside the advisory role, Ungoed-Thomas is a Visiting Senior Research Fellow in the Department of War Studies at King's College London and a Senior Fellow at the Atlantic Council's Scowcroft Centre for Strategy & Security.

Sovereign AI

The appointment comes as debate over sovereign AI expands beyond the location of data and model training. Governments and companies are paying closer attention to how autonomous AI systems are controlled, which systems are trusted, what they are allowed to access and how their actions are checked.

FIOR is focused on that control layer for AI agents. Its AI Agent Enforcement Gateway is designed to identify AI agents, determine what each is authorised to do and enforce those permissions before actions are carried out.

According to FIOR, the system gives AI agents cryptographically verifiable identities and applies policy controls before execution. It can run within an organisation's own infrastructure, including regulated, sovereign and air-gapped environments, rather than relying on an external authority for identity, policy and enforcement.

The business was founded by Chief Executive Officer David Williams, a repeat founder with a background in telecoms, cyber security and fintech. It works with enterprises, governments, technology companies and infrastructure providers deploying autonomous AI systems.

Security focus

FIOR argues that sovereign AI should include not only compute, models and data, but also control of the autonomous software agents operating within those environments. That view reflects wider concern among public sector and corporate technology leaders as AI systems begin taking actions across sensitive data, applications and critical infrastructure.

Williams said Ungoed-Thomas's experience aligns closely with those issues.

"Gemma has spent much of her career dealing with precisely the questions that sovereign AI now raises: how states protect critical capabilities, how technology changes national security, and how we maintain control and resilience as those technologies become more powerful.

"Sovereign AI cannot simply mean owning compute capacity or running a model within national borders. If the identity, permissions and control of the autonomous agents using that infrastructure are dependent upon somebody else, an important element of sovereignty is missing.

"We believe FIOR can provide part of that missing control layer. Gemma will help guide our proposition and ensure that we understand the strategic and operational requirements of governments and major enterprises as this market develops," said David Williams, Founder and Chief Executive Officer of FIOR.

Ungoed-Thomas said the spread of autonomous AI into critical systems raises questions of authority and constraint as much as of infrastructure ownership.

"AI is rapidly becoming part of the critical infrastructure of governments and businesses. As systems become more autonomous, the question of who controls them, how authority is granted and how that authority can be constrained or withdrawn becomes increasingly important.

"Sovereign AI therefore needs to encompass security, resilience and control as well as models, data and compute. FIOR is addressing an important part of that challenge by creating infrastructure through which organisations can identify AI agents and enforce what they are permitted to do.

"FIOR's focus on a deterministic enforcement layer is therefore timely, particularly as organisations look for ways to retain control across increasingly complex AI environments.

"I am pleased to be advising the company as it develops its sovereign AI proposition and considers how its technology can support the security, resilience and control requirements of sovereign AI and major enterprise environments," said Ungoed-Thomas.