Meaningful control over AI
without stifling adoption & opportunity
WE HELP BUSINESS LEADERS RETAIN CONTROL OVER AI
by confronting critical organisational questions, such as
Who actually decides what AI does and how it does it?
What are the real business risks of AI that you need to manage & mitigate?
How well equipped are you to manage AI activity as effectively as human work?
Being accountable for AI is not the same as controlling AI
Your people may be making AI-supported decisions without seeing all the earlier choices that shape what they decide.
For example, when hiring from an AI-generated shortlist, an AI supplier may have set selection criteria, an AI HR system may have filtered CVs as it understood those criteria, and a data team may have defined a "good" candidate.
Of course human oversight matters, but does a human "in-the-loop" at the point of final decision have sufficient control?
We believe meaningful control also requires that person to understand what has already been decided, recognise when it needs challenging, and have a practical way to intervene.
Effective governance connects upstream choices to business responsibility - including beyond your organisational walls.
It helps you establish who can make consequential decisions, what needs to be visible, and when to act, so that AI adoption can progress with clear accountability and workable controls.
AI governance to help your business move forward, safely
We help you manage AI risk responsibly without slowing progress unnecessarily
We help you decide what to look for as AI becomes embedded in your organisation, and what to do with what you see
We help you identify the key AI decisions that determine business outcomes, and extend controls to cover them - including vendors, IT teams, HR and compliance
THERE ARE THREE WAYS WE CAN WORK TOGETHER
Executive briefings & working sessions
Briefings cover a key AI strategy question, typically around AI governance and business control.
Sessions help leadership teams surface key assumptions about AI, clarify expectations and perspectives on what AI will do, and identify who needs to be involved in AI governance
Short, sharp governance reviews
Focused exploration of a current or planned AI-supported activity, such as recruitment. Includes AI activities and responsibilities that shape business outcomes, and how they are controlled.
Develop an organisational governance blueprint for the activity: what is visible and controlled, who can intervene, any further investigation or specialist assurance needed
Ongoing sounding board
Bring an independent perspective to governance as adoption develops. Ongoing discussions can provide challenge and continuity across decisions that would be made separately
The scope and format follow the business question and people who need to address it
Led by Was Rahman
I have spent more than 30 years working across enterprise technology, consulting and founder-led ventures, including leadership roles at Accenture, Infosys and Wipro.
My experience spans the organisational work of choosing, delivering and using technology, as well as the technology itself.
I will shortly submit my PhD at Coventry University on fairness in AI-supported business decisions about people. The research examines how responsibility, differing expectations and upstream choices interact in recruitment and promotion decisions.
That work informs my perspective on organisational AI governance, complementing my experience-based views with a pragmatic adoption of management theory.
Fairness is a particularly interesting and increasingly relevant area of AI governance and control. It is a space in which AI meets the hard-nosed reality of business performance and the most human questions of AI ethics, in an environment of constantly evolving regulation and technical capability.
What AI governance question are you facing?
If you are thinking about how your organisation should direct, oversee or retain control over AI, I would welcome a conversation. Tell me what you are trying to achieve, what is changing and where the uncertainty lies.