What I bring
Independent scrutiny grounded in engineering practice, AI ethics research, international standards, and years of explaining difficult trade-offs to people who have to make decisions about them.
Board and advisory work
AI advisory boards and Non-Executive Director roles
I take selected Non-Executive Director (NED) and Advisory Board roles where AI creates genuinely difficult strategic, ethical, or safety questions — the kind where “responsible innovation” needs to mean something more than a slide deck. I help boards see around corners, challenge assumptions before they calcify, and turn good intentions into decisions that survive contact with reality.
Where I fit
Technology companies, regulated organisations, public-interest institutions, investors, and mission-led ventures whose use of AI has consequences beyond a single product team.
How I work
Candour, curiosity, and a strong preference for governance that actually works. The goal is a board that asks better questions, catches weak signals early, and makes clear decisions under genuine uncertainty.
Board value
From broad principles to board-level grip
Strategic foresight
Map how agentic AI, autonomy, synthetic media, machine welfare, and emerging regulation may alter the organisation’s opportunities, dependencies, and risk envelope.
Governance architecture
Clarify accountability, escalation routes, decision rights, assurance evidence, and the conditions under which a system should pause or stop.
Constructive challenge
Interrogate optimistic assumptions without smothering useful innovation. Surface the questions that tend to vanish somewhere between the model team, the executives, legal counsel, and the board — usually in the gap where everyone assumes someone else is worrying about it.
Trust and legitimacy
Help organisations earn confidence through intelligible commitments, measurable safeguards, and honest communication about uncertainty.
Relevant foundations
Where this comes from
My work spans hands-on machine learning engineering, authorship of Taming the Machine and Safer Agentic AI, leadership in international AI standards, and research into AI governance, agentic AI safety, machine welfare, and bilateral alignment. I am President of the European Responsible AI Office and chair IEEE standards work on transparency, responsible AI practice, and the behaviour of autonomous systems.
Engagement formats
- Non-Executive Director: formal governance, fiduciary oversight, committee work, and sustained independent challenge.
- Advisory Board Member: recurring strategic counsel without a statutory board appointment.
- Board advisor: focused support through a transition, investment, policy decision, or AI governance programme.
- Board and executive briefings: concentrated sessions that build shared understanding before an important decision.
Good fit
- The board faces material decisions about AI strategy, assurance, autonomy, accountability, or public trust.
- The role carries enough access and authority for independent challenge to influence decisions.
- Leadership is willing to examine inconvenient evidence and make governance commitments observable.
Not a fit
- A ceremonial appointment intended chiefly to signal credibility. (I am not a decorative fern.)
- An endorsement of a product or safety claim without the access needed to examine it.
- A role whose commercial interests would compromise independent judgement or conflict with existing duties.
Independence and conflicts
I start every formal appointment with an honest account of duties, decision rights, access to evidence, confidentiality, and potential conflicts. I disclose my own interests and expect the organisation to do the same. Where independent judgement cannot be protected, I will say so and step back.