Skip to main content
About

AI governance is a design problem. That's why I'm here.

I'm James Nicholls. I've spent 20 years making complex systems work for real people — inside UK government, international development, and the private sector. Now I apply that same discipline to AI governance.

The Journey

From neuroscience to AI governance

Not a typical compliance consultancy background — and that's the point.

1999–2000
Neuroscience Research, UCL
Started at University College London researching Long Term Potentiation — the mechanism by which the brain encodes memory. Studying how neurons strengthen connections through repeated stimulation.
"Understanding how humans process information has been the thread through everything I've done since."
2000s
UN & EU Missions, Kosovo
Interviewed 150+ families of missing persons for the UN. Produced films for UN TV including a CNN report. Served as Head of Press for EULEX (EU Rule of Law Mission in Kosovo).
UN21 Award — Kofi Annan, 2005
"Working in post-conflict environments teaches you what happens when systems fail people."
2014–2016
International Development, Nigeria
DFID/FCDO work in Nigeria on public service reform. Met Acting President Yemi Osinbajo. Produced documentaries on youth unemployment and community policing — telling human stories within large institutional reform programmes.
2017–Present
UK Government Digital
Led user research across DfE, Royal Navy, BEIS, MOD, Ministry of Justice, and the Met Office. Brought human-centred design methodology into departments where procurement processes and policy requirements often crowd out the user.
18/18 GDS assessment score — Find Teacher Training (30,000+ users/year)
Led GDS service standard adoption for the Royal Navy. Currently on an active MOD contract working on data services and AI integration.
"Government taught me how compliance actually works — not in policy documents, but in procurement decisions and operational reality."
2024–Present
The Pivot to AI Governance
"I kept seeing the same pattern: organisations adopting AI without the human oversight architecture to make it defensible. The EU AI Act made that problem urgent."
Defensible AI exists because most AI governance advice comes from lawyers or compliance consultants — not from people who understand how humans actually interact with systems. The gap between regulatory intent and operational reality is a design problem, and that's what 20 years of user research has trained me to solve.
Why This Approach Works

Governance that survives contact with reality

01

Human-Centred, Not Just Compliant

Governance frameworks that don't account for how people actually work get ignored. I design governance that organisations can actually implement — because that's what I've spent 20 years doing.

02

Government-Tested

I've seen compliance from the inside — how procurement bodies evaluate vendors, how audit requirements flow through supply chains, what regulators actually look for versus what policy documents say.

03

Practical, Not Theoretical

I don't produce 80-page reports that sit in a drawer. I produce governance frameworks with named owners, clear deadlines, and implementation plans that survive contact with reality.

Credentials

Background & experience

Two decades of making complex systems work for real people — across sectors, borders, and levels of government.

  • Active MOD Contract

    Data services and AI integration for the UK Ministry of Defence

  • 20+ Years in User-Centred Design

    Government, international development, and private sector

  • UK Government Departments

    DfE, Royal Navy, BEIS, Ministry of Justice, Met Office, DFID/FCDO

  • Google Pre-Launch Research

    Product research via INSITUM

  • UN21 Award (2005)

    Presented by UN Secretary-General Kofi Annan

  • UCL Neuroscience Research

    Long Term Potentiation — how the brain encodes memory

  • Published Work

    EULEX book, UN book on missing persons (Issuu, LinkedIn)

Ready to talk?

Book a free 30-minute scoping call. No sales pitch — just a practical assessment of where you stand on AI governance and what to do next.