Research into organisational control over AI
I’m Was Rahman. My research asks what it means for organisations to retain control over AI, and what changes when they use it to support decisions. This site is where I share that work, explore its implications and invite discussion of the questions it raises.
My background spans physics at Oxford, software development and more than 30 years in technology and business. I have also been a Visiting Researcher at Bonn and a Research Fellow in the Cambridge–Bonn Desirable Digitalisation programme. I bring these different perspectives to my research, looking at where they reinforce each other, where they conflict and what we can learn by considering them together.
My doctoral research at Coventry examines fairness in AI-supported recruitment and promotion. It provides a foundation for my wider inquiry into organisational control: not only how things work in practice, but how we understand and explain them.
What does retaining control involve?
Keeping a person in charge does not necessarily give them control. My research looks at five connected questions about the choices organisations make, how they work, and what changes when they use AI.
My PhD explores these questions through fairness in recruitment and promotion, providing a foundation for my wider research into organisational control over AI.
By the time someone reviews an AI recommendation, earlier choices may already have limited their options. What the system is designed to achieve, which information it uses and where its thresholds are set all help shape the eventual decision.
Those choices may reflect reasonable business priorities, such as reducing costs or working faster. But the person making the final decision needs to understand which priorities have already been set and how these limit their options. They also need a way to question earlier choices when the consequences suggest those choices should change.
An organisation can sincerely commit to a principle while its targets, budgets and everyday practices pull people in another direction. A new policy does not, by itself, change those pressures or the established ways people work.
Management theory helps explain why these gaps persist, without assuming that leaders are uncaring or that their commitments are dishonest. My research draws on these explanations to understand what happens when AI enters the picture: which practices it carries forward, and whether governance changes what happens or simply shows that a process has been followed.
AI can help organisations work faster and more consistently. But deciding what to do is not always the same as applying a rule: people may need to interpret the circumstances, weigh competing concerns or recognise when an exception matters.
Human judgement is not automatically better. The question is what changes when we turn that judgement into something a computer can do. My research examines what this translation preserves, what it leaves out, and what people still need to be able to question or decide.
A manager may be responsible for a decision without being able to change the system that helped produce it. They may not know how it works, have access to the people who shaped it, or have the authority to ask for changes.
Naming someone as accountable does not resolve that gap. My research examines how responsibility connects to the practical means of exercising control, including when those means sit with another team or an outside supplier. Spotting a problem is only useful if there is a route to getting it addressed.
Before AI can use information, it has to be represented in a form the system can work with. Turning an assessment of a person into a category or score can make comparisons easier, but it can also leave out context or change the meaning of the original assessment.
That matters when we use the results to judge whether a decision is good or a process is working as intended. A measure may accurately describe what was counted without telling us everything we need to know. My research examines what is lost or changed in that translation, and what we can reasonably conclude from the data that remains.
What the research is based on
My PhD at Coventry University examines how organisations deal with fairness when AI supports decisions about people, particularly recruitment and promotion. It draws on 30 in-depth interviews with people working in HR, business, technology and academia.
The interviews explored what fairness meant to people, how they tried to put it into practice, and what happened when AI became involved. Three recurring issues help explain why this is not simply a matter of choosing a fairness measure and applying it:
- People meant different things by fairness. Agreement that fairness mattered did not necessarily mean agreement about what a fair decision would look like.
- Fairness competed with other priorities. Cost, speed and commercial goals shaped what people could do, alongside the resources and information available to them.
- Responsibility and influence did not line up neatly. Those expected to oversee people decisions were not necessarily connected to everyone shaping how AI supported them.
Management theory helps explain why these difficulties persist, including when people sincerely want to act fairly. The thesis uses those explanations to examine how AI can carry existing organisational choices into systems, making some harder to see or reconsider.
What the thesis adds
The work develops two contributions to management theory. Both address problems that become important when different people shape a decision at different stages.
The first builds on agency theory, which examines relationships in which one party acts on another’s behalf. It considers how choices made by developers, suppliers and others can influence a decision without passing through the judgement of the person making it, and what organisations need to govern that influence.
The second builds on boundary object theory, which explains how people can work together using shared tools or ideas even when they understand them differently. It examines when those differences matter to the outcome and proposes a framework for making different understandings of fairness visible, so organisations can decide what to do about them.
These contributions combine interview evidence with theoretical analysis. Further research is needed to investigate upstream practices directly and to develop and test the proposed approaches in organisations.
About me
I studied physics at Oxford, specialising in atomic and quantum physics in my final year, before completing a master’s in Information Systems Development and Management. My early career involved developing business software and helping others use software engineering methods and tools. I went on to leadership roles at Accenture, Infosys and Wipro, as well as building and running businesses.
That background shapes how I approach questions. I look at them from several perspectives, exploring where those perspectives reinforce each other, where they conflict, and where bringing them together reveals something we might otherwise miss. Each has something important to contribute, but none automatically takes precedence. Sometimes the result is a new understanding of something we thought we already knew.
My academic work brings these questions into conversation with management, philosophy, law and computer science. I was a Visiting Researcher at Bonn’s Center for Science & Thought from 2023 to 2025, and a Research Fellow in the Desirable Digitalisation programme, based across Cambridge and Bonn, from 2023 to 2024. My current academic home is Coventry University’s Faculty of Business and Law.
Research and conversation
I welcome conversations with researchers and business leaders who want to explore these questions, challenge an argument or compare different ways of understanding a problem. I’m interested in collaborative research, speaking and discussion, and visiting academic opportunities through which I can continue developing this work.
Alongside my research, I run AI Prescience, a small professional practice where I apply this work to help organisations achieve their business goals. This site focuses on the research, the ideas it develops and the questions still open to debate.