Digital self-services have become a central part of digital government, enabling citizens to apply for benefits, submit documentation, and interact with public authorities online. While these systems aim to improve efficiency and accessibility, they also shift administrative work, such as interpreting rules, navigating procedures, and demonstrating eligibility, from public officials to citizens. As a result, interacting with digital public services can generate administrative burdens for citizens in the form of learning, compliance, and psychological costs. Although administrative burden theory offers a useful lens for understanding such experiences, research and public organizations still lack practical approaches for identifying where these burdens arise within digital service delivery. Existing research often focuses on policy design, system usability, or citizen outcomes, providing limited guidance on how burdens become visible in concrete digital service encounters and how they can be traced to the socio-technical arrangements that shape them. This paper develops an empirically grounded, practice-based conceptual framework for identifying administrative burdens in digital self-services. The framework links citizens’ experienced burdens to policy design, digital infrastructures, and organizational practices and operationalizes this perspective through a diagnostic tool for analyzing digital services. Drawing on iterative engagement with qualitative material from research of digital welfare services in Norway, the paper further demonstrates the analytical application of the framework through examples from the Norwegian Labour and Welfare Administration (NAV). The framework contributes to digital government research by providing a structured approach for diagnosing administrative burdens and identifying potential intervention points in digital public services.
Digital sovereignty—the capacity of states to exercise control over digital infrastructures, data, platforms, and technological development—has become a critical concern in the era of Artificial Intelligence (AI)-driven digital transformation. As governments increasingly rely on global cloud providers, proprietary AI systems, and transnational data ecosystems, questions of autonomy, accountability, and long-term strategic capability intensify. Existing literature, however, offers limited empirical insight into how digital sovereignty is understood and operationalised in the Global South, particularly within AI-enabled public sector contexts. This study addresses this gap by developing and empirically grounding a multi-dimensional framework of digital sovereignty in Kenya’s public sector. Employing a qualitative case study approach, the research draws on fourteen semi-structured interviews with government, policy, and industry stakeholders. The findings reveal that digital sovereignty is negotiated across operational, strategic, and epistemic dimensions, shaped by infrastructural dependency, fragmented governance, and trade-offs between efficiency and control, with implications for trust and service delivery.
For over a century, the public sector has used information technology to enable and streamline administrative tasks. Today, welfare benefit administration in Scandinavia predominantly occurs through a combination of automation and digital self-service, both of which rely heavily on national data registries. This has led to increased government efficiency in many areas, and improved service quality for most citizens. However, automation and digital self-service also impose several demands on citizens regarding IT access, skills and competencies, as well as their personal characteristics and life history. In this paper, we explore and illustrate these demands through a series of vignettes based on in-depth qualitative studies of citizens and frontline employees in Scandinavia. We conclude by classifying the various types of demands and discussing them in relation to existing research and current practices for public service provision.
AI-enabled mass deliberation is a new attempt to revive deliberative democracy with digital technology. AI applications such as summarisation and automated moderation provide opportunities to scale up deliberative processes that currently are constrained to discussions in small groups. Notwithstanding, the AI push underestimates the difficulties in sustainably integrating digital technologies (including generative AI) into sedimented organisational and normative practices of decision-making. This paper aims to understand the gap between institutionalization of AI-enabled deliberation as a conceptual notion and its effectuation in practice. For this, we collected and synthesised well-established conditions and requirements for institutionalisation through co-creation sessions with experts. Thereafter, we asked practitioners to assess these conditions and requirements on importance, contribution to institutionalisation, feasibility, and easiness of implementation. Our research identifies five tensions in practice that contribute to the gap between the conceptual understanding of institutionalisation and practitioners’ perceptions of whether it is achievable in practice. These tensions need to be overcome for institutionalisation to be accomplished.
There are, in most countries, many digital systems used for communication between home and school, concerning administration, planning, implementation, and evaluation of education and students’ performance. This complexity might both be an asset and a burden to administrate for both teachers, legal guardians, and students. Education is in most countries compulsory and public funded thus digitalization in schools, and not at least home-school relations, is a form of digital government. This paper contextualises the communication between home and school as digital government through a mapping literature review with focus on their methods, analytical frameworks and cases to identify mediating components in the communication and identify research gaps. Pedagogical or didactical research was excluded to focus on education as a public service, and the digital communication between home and school as a form of digital government. The scope is explorative and aims to serve as a foundation on which to base further empirical studies in the area. The findings show that the included studies (n = 19) are predominantly qualitative and within the field of education and teaching, both considering applied theoretical models and outlets. By considering the home-school communication as a form of digital impact mediator, drawing on Helsper’s corresponding fields model [8], we structure the results and suggest both theoretically and empirically generated openings for further research. By studying digital communication between home and school through a digital government lens we draw the conclusion that further research in the area is needed to explore the factors that connect digital and offline exclusion in a school context.
Although public planning instruments are widely institutionalized, they still lack systematic mechanisms capable of objectively assessing their completeness and alignment. As a result, plans often exhibit gaps and inconsistencies that hinder the translation of strategic intentions into effective actions. This article addresses this limitation by proposing the UPICEF-A framework, designed to assess the completeness and alignment among instruments such as Strategic Planning, the Goal Plan, the Multi-Year Plan, and budgetary instruments. Methodologically, the study combines a conceptual analysis of planning instruments with the development of an evaluation model that structures objectives, goals, initiatives, and indicators into verifiable relationships. The framework enables the identification of gaps, overlaps, and misalignments, as well as the assessment of the completeness of planning artifacts. Results indicate that the consistency observed in formal documents does not fully hold when analyzed relationally, revealing weaknesses in the articulation between strategic and operational levels. By providing an objective mechanism for evaluating these relationships, UPICEF-A contributes to shifting public planning analysis from a purely formal perspective to a structured approach focused on completeness and alignment, supporting the development of more coherent and effective planning processes.
Artificial intelligence (AI) is rapidly reshaping public administration by influencing how governments design services, manage data, and exercise regulatory authority. National AI strategies have become essential tools for governments to articulate their ambitions and define their role in governing AI, yet the policy landscape has evolved significantly since their adoption. This paper examines how national AI strategies define the role of public administration in AI governance. Drawing from concepts in AI governance, digital government, and policy design, we propose a role-based analytical framework that conceptualizes public administration as AI adopter, regulator, and orchestrator of AI ecosystems. We apply the framework in a qualitative comparative policy analysis of national AI strategies in Germany, the Netherlands, and the United Kingdom. Using a deductive-inductive qualitative content analysis, we reconstruct country-specific role profiles and compare strategic emphases across the three cases. Our findings reveal that all three strategies foreground the adopter role, with extensive public-sector use cases and capacity-building measures, while systematic scaling beyond pilots is made explicit only in the UK’s “scan, pilot, scale” approach. Regulatory framings are more prominent and tightly coupled to EU-level developments in Germany and the Netherlands, whereas the UK emphasizes innovation and competitiveness with fewer detailed legal commitments. Orchestration arrangements likewise diverge, ranging from partnership-oriented ecosystem coordination in the Netherlands and the UK to more hierarchical steering in Germany. We argue that a role-based lens helps to uncover how early strategic choices position public administrations within emerging AI ecosystems and to identify misalignments between these framings and the evolving European AI governance landscape, thereby providing an empirically grounded lens to assess whether existing strategies remain fit for purpose.
Citizen-centricity has become a dominant paradigm in digital government, often framed either as a novel transformation of public administration or as an incremental improvement in service delivery. Yet little research has examined how this paradigm reshapes the production of knowledge about citizens within digital governance. This study addresses this gap by analyzing citizen-centricity as an epistemic transformation rather than solely a managerial or technological innovation. Drawing on fourteen semi-structured interviews with officials involved in a digital government initiative in Belgian public administrations, this article mobilizes Michel Foucault’s archaeology of knowledge to examine how contemporary governance constructs knowledge about citizens. The findings reveal three interrelated dynamics: (i) the proliferation of user multiplicities as fragmented and context-dependent categories, (ii) the conceptualization of users as variable configurations of attributes, and (iii) the co-production of the user through interactions between citizens and public administrations. These dynamics indicate a shift away from the modern episteme described by Foucault, which centers knowledge on the human subject as an abstract, stable, and universal unit. Taken together, the findings contribute to eGovernment research by reframing citizen-centricity as an indicator of epistemic transformation in the production of knowledge about citizens, while also highlighting implications for practitioners and suggesting research directions to address them.
Dashboards are increasingly deployed in the public sector to support transparency and data-driven decision-making. However, existing practices of dashboard design often lack a holistic, end-to-end approach that integrates public-value considerations throughout the dashboard development lifecycle. This study develops a research-grounded framework using the Design Science Research Methodology (DSRM). We synthesize from a literature review and consolidate methods, techniques, design principles, development guidelines, existing frameworks, and implementation cases. These insights are integrated into a structured dashboard design lifecycle comprising multiple phases and sub-phases that guide the transformation of raw data into interactive dashboards. The framework is operationalized through the ErLE dashboard, which visualizes e-participation offers in German municipalities.
In this paper I examine how Norwegian political parties used social media in the 2025 parliamentary election campaign, with particular focus on Facebook, as Facebook remains the largest social medium for political communication. Using genre theory as analytical lens, and the public sphere as theoretical perspective, I examine how political campaigning in social media has evolved from earlier elections to the current (2025) election and discuss if Norwegian political communication is moving more towards an affective public sphere and thereby adopting the rhetorical devices of populism.
Despite the growing use of AI in the public sector, there is limited empirical evidence on its implementation. Prior research on government readiness for AI has predominantly focused on the adoption stage from the viewpoint of IT or public managers, often yielding high-level insights and one-sided narratives. AI implementation goes beyond adoption as it involves shaping, integrating, and routinizing AI applications that are often overlooked in discussions about AI adoption. This study examines how government readiness for predictive AI implementation involving multiple stakeholders can be evaluated. It develops a framework for evaluating government readiness for predictive AI implementation by analyzing three Indonesian government use cases and one Dutch government case. The framework covers the socio-technical factors of AI and enables more nuanced analysis by involving perspectives of: 1) those of the stakeholders responsible for the development, management, and governance of AI innovation (which, in short, we call AI developers), and 2) those of the AI users. From developers’ perspectives, the framework examines readiness in terms of resource availability, the organization’s capability to change, and governance, while from users’ perspectives, readiness is evaluated based on the perceived performance and impact of the deployed AI systems. The framework also indicates the influence of AI characteristics and individual readiness on organizational-level implementation. The scientific contribution is to extend the AI readiness model in government beyond adoption, addressing multiple stakeholders’ perspectives on implementation. This framework is expected to guide public organizations in reflecting on their AI implementation and make iterative improvements to ensure responsible AI.
Every five years, a global stocktake (GST) is conducted to evaluate progress towards the goals in the Paris Agreement. The GST involves processing large quantities of information produced by a diverse range of stakeholders. Scholars have suggested exploring the use of Generative AI (GenAI) technologies such as large language models (LLMs) to support the GST. To examine this proposal, we conduct a scoping review to identify challenges in the GST and map these challenges to a framework of GenAI modes of operations, and literature on trustworthy AI. We propose three problem-solution mappings (information challenge, effectiveness, GST outcomes) where GenAI can support the GST process. Then, we present important considerations that need to be addressed when using AI in this context and propose scenario building as a path for future research. By doing so, this paper links literature on climate negotiations and AI, and draws on principles of trustworthy AI to guard against inaccuracy, bias, and inadequate human oversight.
The Oversight of Artificial Intelligence is emerging as a complex challenge. To tackle this complexity, regulators are increasingly required to collaborate on AI oversight in networks. Yet, little is known about how the coordination of AI oversight must be organized in these arrangements. This paper identifies factors that influence the coordination of AI oversight in regulatory networks, focusing on the Dutch Samenwerkingsplatform Digitale Toezichthouders (SDT). We conducted semi–structured expert interviews with members of AP, DCA (Directie Coördinatie Algoritmes), who are tasked with facilitating collaboration and coordination of AI Oversight in the SDT network. We identify overarching themes that influence coordination in three interlocking core AI oversight activities: risk monitoring; joint evaluation of AI systems; and joint advice for legal compliance. Our findings show that AI oversight has wicked characteristics whose resolution is dependent on the critical functions of knowledge integration. To strengthen knowledge integration, we recommend that regulators work towards system-level assessment methods for AI systems that readily allow them to combine their different perspectives. We believe that such assessment practices should embrace the principles of knowledge co-production. Our descriptive approach helps inform coordination of AI oversight in the Netherlands and elsewhere.
Data portability, initially established to guarantee individuals’ self-determination over personal information, has evolved into a critical mechanism for fostering a fair data economy. However, its practical implementation often faces significant challenges due to technical barriers and conflicting stakeholder interests. This study analyzes the success and failure factors of data portability across various global cases—including UK Open Banking, Korea’s MyData service, and US vehicle data access laws(Right to Repair)—to identify the essential requirements for institutionalizing these rights. Based on these analyses, we propose the ‘Data Portability Success Index ( S_DP )’ equation, which quantifies success through two primary variables: the reduction in switching costs via technical standardization ( C_s ) and the ratio of trusted intermediaries ( R_ti ) within the infrastructure. Our findings indicate that technical standardization and trust-based governance are more decisive for the effectiveness of data portability than mere legal codification. Consequently, this paper suggests that the government must fulfill dual roles as a ‘Technical Standard Architect’ and a ‘Trust Governance Builder’ to minimize market uncertainty and promote a robust data ecosystem. This research provides a theoretical and quantitative framework for future policy prioritization and empirical analysis across diverse industries.
The European regulatory framework for data spaces, established by the Data Governance Act (DGA) and the Data Act (DA), introduces various legal requirements surrounding the development of the data economy across different sectors. Complying with these legal requirements while also ensuring an economically sustainable business model is particularly challenging in sectors such as the electronics and the textile industries, which often involve complex international value chains including several participants of different sizes and market power, many of which may not widely employ technological solutions from the digital age. Focussing on the electronics and textile industries as general examples, the paper at hand analyses the European regulatory framework for data spaces and investigates how state-of-the-art data space technology can provide (at least partial) solutions to emerging legal challenges.
Data preparation plays a central role in shaping how administrative categories become available for computational use, yet its consequences often remain difficult to trace once algorithmic systems are operational. This paper conceptualises large-scale dataset annotation as infrastructural and anticipatory work through which administrative distinctions are prepared for later algorithmic applications. Drawing on sociotechnical perspectives on data, infrastructure, and anticipatory governance, the paper examines how large-scale annotation practices contribute to different future orientations: future suspicion, future identifiability, and future knowledge and representation. Using illustrative analytical vignettes from public authority contexts, the analysis explores how annotation establishes conditions for later forms of scrutiny, recognition, and knowledge production while separating where classifications are established from where their consequences later emerge. By shifting analytical attention from algorithmic outputs and model performance to annotation, the paper highlights data preparation as a constitutive site of governance through which what becomes visible, actionable, and governable is organised. In doing so, the paper develops a temporal account of how classifications travel across organisational contexts and acquire consequence at a distance from the settings in which they were originally produced. Positioned within critical data and AI studies, the paper contributes by articulating the classificatory and infrastructural work through which the conditions for later algorithmic action are assembled and by examining the implications for accountability related to public authorities’ deployment of AI systems. .
While citizen participation provides essential local expertise for effective policies, many initiatives struggle to move beyond consultation. Progressing toward policy implementation is often hindered by limited resources and technical complexity. This study investigates the potential of AI representatives to bridge this gap. Using exploratory interviews with an embedded vignette, we explore how 52 citizens and public servants in a German municipality perceive the legitimacy of AI agents acting as digital representatives during internal policy formulation meetings. Our findings provide empirical evidence that AI representatives are perceived as a potentially legitimate mechanism to ensure that citizens’ voices persist throughout the policy cycle. We identify the specific conditions and rationales under which AI representatives are seen as a viable compromise to shift participation from simple consultation toward a more structural role in governance.
Digital platforms increasingly use gamification mechanisms such as points, badges, and leaderboards to boost participation in democratic deliberation. Yet deliberation is a multi-stage process, each with distinct cognitive demands and normative priorities, including equality, inclusion, plurality, authenticity, and reflection. Applying a uniform points-badges-leaderboards (PBL) gamification layer can therefore create design-process mismatches, where incentives that are functional in one stage become counterproductive in another. Based on a stage-by-stage analysis of common deliberative modules, the paper maps how standardized gamification can distort participant acquisition, group building, learning, problem definition, open discussion, self-reflection, decision-making, and output generation and validation - for example by amplifying homophilous recruitment, privileging speed over comprehension, or threatening plurality through quantity biases and gatekeeping effects in problem framing. The paper then introduces a normative process-sensitive alignment framework and a “stage-fit test" that link each deliberative module to suitable extrinsic- or intrinsic-leaning mechanisms, associated risks and affordances, and concrete design safeguards. The framework is theory-driven and requires empirical validation in concrete deployments. Finally, the paper derives policy implications for treating gamification as a form of governance, proposing stage-bound design standards, quality-first evaluation criteria, and transparency obligations for public-sector deliberation platforms. This paper offers a normative, process-sensitive framework and stage-fit test.
As governments across Europe expand citizen access to electronic health records, access alone does not guarantee comprehension. This study asks whether AI-mediated presentation of the same persistent government health data produces better outcomes than a conventional EHR interface, and whether this matters differently for different user groups. A within-subject experiment with 36 participants across three groups (clinicians, health AI/IT professionals, and non-expert citizens), was conducted and each evaluated a mock EHR and an AI-mediated presentation of the same synthetic patient records. AI-mediated presentation produced statistically significant advantages in usability, cognitive load, and task accuracy. Medium-to-large effect sizes were also reported, along with strong Bayesian support for the two primary outcomes. Citizens reported the largest gains. Qualitative findings also revealed that trust in AI emerged as a distinct dimension. Presentation format appears to be a variable that may affect who can effectively use citizen-facing government health data.
Responsible AI has become a widely adopted reference point in contemporary AI governance, reflected in a growing number of ethical guidelines, regulatory initiatives, and organizational governance frameworks. However, the meaning of “responsibility” often shifts as governance moves across institutional contexts, creating translation gaps between normative principles, regulatory frameworks, organizational practices, and citizen experiences. As a result, Responsible AI governance frequently appears coherent at the policy level while producing tensions or legitimacy concerns in practice. This study aims to explain how responsibility varies across governance levels and how such variation creates cross-level tensions in Responsible AI governance. Based on a systematic literature review and typology development, we synthesize the Responsible AI literature to develop a typology of Responsible AI governance across four governance levels - super-macro, macro, meso, and micro - each characterized by distinct institutional logics, legitimacy criteria, and modes of socio-technical enactment. The analysis shows that tensions in Responsible AI governance arise not only from implementation gaps but also from structural differences in how responsibility is defined, operationalized, and evaluated across governance levels. The study contributes to Information Systems and AI governance research by conceptualizing responsibility as a relational, multi-level governance construct, advancing understanding of persistent principle-to-practice gaps, and offering a diagnostic tool for policymakers, organizations, and researchers designing or auditing Responsible AI governance frameworks.
