Document Overview
This is a peer-reviewed research article by Illugi Torfason Hjaltalin and Hallur Thor Sigurdarson, published in Government Information Quarterly (2024). The authors are affiliated with Reykjavik University. The article examines how national AI strategies define and position artificial intelligence applications within the public sector.
Research Question and Scope
The authors state that the study investigates how governments define and position AI in the public sector. The overarching research question asks what the discursive positions in national strategies are concerning AI applications for important government functions, and how the texts convey democratic and managerial public values differently in this context. The analysis covers 28 national and cross-national AI strategy texts from nine major economies.
Theoretical Framing
The article uses public value theory, drawing on Moore (1995). It describes public value as a desired outcome of government initiatives, comparable to how private organizations create economic value. The authors also refer to a three-part classification of AI in government: policymaking, public service delivery, and internal management, as proposed by Misuraca and Viscusi (2013).
Methodology
The authors adopted combined qualitative methods. The primary method was discursive positioning (Wood and Kroger, 2000), a form of discourse analysis. The team manually extracted references to the public sector, coded examples of AI application by sector, and identified dominant discourses and how they conveyed public values.
Key Findings
The article identifies three dominant discourses: empowerment through information, enhanced administrative practices, and improved service delivery. Enhanced administrative practices positioned AI as a tool to optimise internal processes, resource allocation and organisational management. Improved service delivery placed public services at the centre, while empowerment through information framed AI as enhancing citizens' service experiences. The authors report that discourses on the policymaking function received limited attention. They also find that efficiency and service delivery dominate the discourse, while citizen engagement is underemphasised. The public values envisioned include efficiency, agility and inclusion.
Discussion and Implications
The authors argue that viewing AI only as automation may constrain other innovative solutions for public value. They state that policymakers must strike a balance, ensuring AI aligns with broader societal outcomes while addressing democratic imperatives. The study notes that stakeholder engagement, transparency and accountability are widely viewed by earlier scholars as critical to aligning AI strategies with public values.
Why It Matters for AI Policy in Government
According to the article, the study aims to extend public value research in e-government and to add to the growing body of work on public sector AI. Its stated audience includes researchers in public administration and policymakers designing national AI strategies.