What did you study?
I studied how principles such as fairness, transparency, and accountability can be translated into ethical requirements: concrete expectations that guide how AI systems are developed and managed.
What were the results of your study or what is its main finding?
Ethical requirements need ongoing attention; a one-time checklist is not enough.
The dissertation develops practical approaches for this. The Ethical Requirements Canvas helps organisations evaluate ethical requirements in relation to stakeholder needs, risks, costs, benefits, and organisational value. The Ethical Requirements Stack shows how high-level ethical principles can be translated into concrete development requirements and tasks, while a governance extension supports responsible data management and accountability.
Together, these inform Adaptive Ethical Requirements Frameworks, which support reviewing and updating ethical requirements as AI systems change. AI tools can assist, but human judgement remains essential.
How can the results be applied? What new insights did the research contribute to the topic?
Software organisations can use these approaches to turn ethical concerns into development requirements and consider business value such as risk reduction, compliance and user trust.
For example, imagine an AI fitness app that first uses basic exercise data but later begins using location data. A privacy requirement that was sufficient at the start may no longer be enough. The approaches developed in the dissertation can help organisations translate privacy into concrete development requirements and review them as the system changes. This can reduce emerging risks while supporting compliance, user trust and organisational value.
The research connects ethical responsibility with software development and organisational value.
Public examination
MSc Mamia Ori-otse Agbese’s doctoral dissertation, Towards Trustworthy AI in Software Engineering: Operationalizing Ethical Requirements through Adaptive Frameworks to Support Business Value, will be publicly examined at the University of Jyväskylä, Faculty of Information Technology, on Friday 30 October 2026 at 12:00 in room H320, Historica.
The opponent is Professor Sami Hyrynsalmi, and the custos is Professor of Practice Tapio Frantti.
Livestream: [https://r.jyu.fi/dissertation-agbese-301026]
Further information:
Mamia Ori-otse Agbese
Mamia.o.Agbese@student.jyu.fi
