Research
The Risks of AI to Strategy
An article-based PhD on how generative AI reshapes strategic decision-making, and how to keep it responsible.
My PhD, “Safeguarding Responsible Strategic Decision-Making in the Age of Generative AI”, asks how generative AI reshapes strategic decisions by altering judgement, cognitive authority and responsibility, and how those changes can be governed responsibly in practice.
The programme
The thesis is article-based and sits at Sheffield Business School, Sheffield Hallam University, under the direction of Professor Seun Kolade. It uses New Institutionalism to follow generative AI across three levels, and maps cognitive risk through three stages of strategic cognition: search, representation and aggregation.
Across the articles, one finding recurs: responsibility and governance attach to what organisations can see and control, whereas GenAI’s influence and its risks arise earlier and elsewhere, in a technology controlled by others.
DIGEST 2026: Best Paper Award
Best Paper Award, Track 3: Inclusion, Equity and Digital Justice
My paper “Phronetic Leadership: A Virtue Ethics Approach to Strategic Decision-Making with Generative AI” was accepted at the DIGEST 2026 conference (Digital Transformation in Industry, Government, Education, and Society). I presented the paper, spoke on the conference panel and received the Best Paper Award for the track.
Risks: more insight, less foresight
More Insight, Less Foresight: Generative AI and the Narrowing of Strategic Decision-Making
A systematic-narrative review of 102 studies that identifies 25 risks along two dimensions: where they sit (human-cognitive, technical, institutional, organisational) and whether they are inherited, amplified or novel. Eighteen of the 25 sit in the cognitive and technical loci, upstream of choice.
Governance: what firms disclose and what they practise
Architectural Decoupling
An analysis of 451 annual reports from 41 listed firms across Asia, Europe, the US and Africa (2015–2025), together with 47 expert interviews. Firms build governance into the architecture of use, while the risks originate in model architecture they do not control.
Bounded Openness: Governing Transparency and Inclusion in AI-Mediated Strategy
212 disclosures of five frontier-AI firms (2012–2025) show how openness in strategy is bounded. The paper proposes a governance principle: open by default, close by exception, reopen by date.
ASTRA: Auditability, Scrutiny, Traceability, Resilience, Awareness
A governance framework validated through a two-round modified Delphi with 19 experts (consensus on 32 of 39 items). The ASTRA Record is a one-page artefact that travels with a strategic decision.
Judgement: phronetic leadership
Phronetic Leadership: A Virtue Ethics Approach to Strategic Decision-Making with Generative AI
Three disciplines for leaders working with GenAI: own the judgement, pause before commitment, stand behind the choice.
Authors or Validators? Practical Wisdom and the Exercise of Agency in AI-Assisted Strategic Work
The journal development of the DIGEST paper: when AI drafts the analysis, do leaders remain the authors of strategy, or become its validators?
From Tool to Authority
How claims about AI shift from describing a tool to asserting its authority in organisational decisions.
Context: education and the Global South
Leading Under AI Deprivation: Generative AI and Strategic Decision-Making in the Global South
Fifteen expert interviews produce a five-dimensional taxonomy of AI capability deprivation and reveal a verification-burden paradox: where capability is scarce, checking AI output costs most.
AI and Primary Education Governance: The Dynamics of Strategic Narrowing
Applies the narrowing lens to primary education policy and identifies three forms of narrowing: input, representation and legitimation.
Work in progress is shown with its status only. Links to full texts will be added as each paper is published.