AI Landscape Digest

Generated on: October 15, 2025

📚 Research Highlights

Recent Developments in Digital Public Infrastructure, Data Governance, and AI in Public Services

Digital Public Infrastructure Framework and Development

The World Bank has recently published a comprehensive strategic framework titled "Digital Public Infrastructure and Development: A World Bank Group Approach" [1]. This white paper defines DPI as "foundational, digital building blocks designed for public benefit" that serve as the core infrastructure layer supporting various services [1]. The document emphasizes DPI systems as interoperable, modular, and reusable systems incorporating open standards to ensure seamless integration across platforms and services [1].

In a significant development, the World Bank has launched a new cross-sector Digital Public Infrastructure (DPI) Program to broaden its support for DPI implementation and use [2]. The program, announced on October 7, 2025, aims to help nations design, implement, and maintain DPI that is "fit-for-purpose, context-specific, and grounded in inclusion, trust and impact" [2]. According to the World Bank's report, "Between 2022 and 2024, Digital Public Infrastructure evolved from a nascent concept into a central pillar of the digital agenda" [2].

The OECD has also released a paper exploring digital public infrastructure as "shared digital systems that are secure and interoperable and that can support the inclusive delivery of and access to public and private services across society" [3]. The paper identifies key components of DPI including digital identity, payments, data sharing, digital post, and core government data registries, and highlights the pivotal role governments play in designing, developing, and managing DPI [3].

AI Governance and Public Administration

The OECD has published a new report, "Governing with Artificial Intelligence: The State of Play and Way Forward in Core Government Functions," which analyzes 200 real-world examples of how governments are using AI across 11 core government functions [4]. Approved and declassified by the Public Governance Committee on September 5, 2025, the report examines how AI is being adopted across government operations, identifies benefits and risks, and proposes a framework for ensuring trustworthy use of AI in public administration [4].

The report categorizes AI benefits in government into four main areas: automation of repetitive processes, strengthened decision-making and forecasting capabilities, improved accountability through anomaly detection, and new opportunities for citizens and businesses [4]. However, it also emphasizes significant risks including biased algorithms, insufficient transparency, over-reliance on AI, and potential public service workforce displacement [4].

To address these issues, the OECD proposes a framework based on three pillars [4]:

Enablers: Governance structures, digital infrastructure, data management, funding, and workforce skills
Guardrails: Clear rules, accountability measures, transparency requirements, and oversight bodies
Engagement: Mechanisms with citizens, civil society, and businesses for designing user-centered AI systems

AI and Data Governance Innovations

Recent research by Batool et al. (2025) has conducted a systematic literature review of AI governance, analyzing 28 research papers to categorize key governance elements under five levels [5]. The study highlights the need for effective governance mechanisms as AI technologies advance at an unprecedented pace, with a focus on systematically analyzing implementation approaches [5].

A newly published paper in the journal "AI and Ethics" addresses how artificial intelligence is transforming various sectors while introducing different types of risks that require identification, assessment, and mitigation [5]. The researchers note that while various AI governance frameworks have been released recently by governments, organizations, and companies, it remains challenging for AI stakeholders to identify the most suitable framework for their AI system [5].

A significant data-centric approach to AI governance is emerging, as highlighted in a paper titled "Data-Centric AI Governance: Addressing the Limitations of Model-Focused Policies" [6]. The research illustrates the importance of considering dataset size and content as essential factors in assessing the risks posed by AI models. The authors argue that current regulations on powerful AI capabilities are narrowly focused on "foundation" or "frontier" models, while policy debates often fail to consider the data used with these models, despite the clear link between data and model performance [6].

Public Administration Applications and Challenges

A recently published study in the journal "Administrations" explores the application of artificial intelligence in public administration, examining its potential to enhance efficiency, sustainability, and resilience in government actions [7]. The research develops a theoretical framework to assess the relationship among AI integration, governance improvements, and economic benefits [7]. The analysis finds that AI does not simply restrict or enhance discretion but redistributes it across institutional levels, potentially strengthening managerial oversight while enhancing decision-making consistency [8].

Another study published in February 2025 in the journal "AI" examined the adoption of artificial intelligence in public administration by analyzing 3,149 documents from the Scopus database to identify the top 200 most-cited articles [9]. The research found that automation of repetitive processes, tailoring of services to individual needs, strengthened decision-making capabilities, and improved accountability are key benefits of AI implementation in government [9].

Frontier AI Models and Data Governance

A June 2025 update to a paper titled "Towards Data Governance of Frontier AI Models" introduces an approach called "frontier data governance," which opens up new avenues for monitoring and mitigating risks from advanced AI models [10]. The paper provides a brief overview of 15 technical mechanisms and introduces five unexplored central recommendations: canary tokens, data filtering, reporting requirements, data security, and know-your-customer regulation [10].

The researchers emphasize how policymakers can use unique mechanisms within this approach to prevent the acquisition of specific dangerous capabilities, whether caused by malicious actors or potential misalignment [10]. The framework provides specific, technical mechanisms for enforcement, addressing a key limitation of current AI governance proposals [10].

Emerging Global Trends in AI and DPI Governance

Data for Policy, in partnership with the University of Cambridge Centre for Science and Policy (CSaP) and Cambridge University Press, has announced a new Global Fellowship Programme focusing on Digital Public Infrastructure, Algorithmic Governance, and GovTech [11]. This 12-month executive fellowship combines flexible online learning with an immersive residency at the University of Cambridge [11].

In Asia, significant AI governance and innovation developments are accelerating [11]. Japan has passed an AI governance bill balancing corporate accountability with innovation incentives, while South Korea has unveiled reforms to strengthen its digital economy alongside ethical AI development [11]. China has designated a landmark generative AI legal case as "typical," underscoring AI accountability, and announced the establishment of 10 national data zones to drive AI research and digital economic growth [11].

University Research on AI in Public Administration

Recent scholarship has increasingly emphasized the transformative potential of artificial intelligence in public administration, highlighting the role of various stakeholders [7]. A special issue on Artificial Intelligence and Public Administration: Actors, Governance, and Policy explores the need for an analytical framework of AI in the public sector based on three levels of public administration: macro, meso, and micro [12].

The article by Ruvalcaba-Gomez, "Systematic and axiological capacities in artificial intelligence applied in the public sector," explores the micro-level of public administrations with a comprehensive attention to competencies of public employees using AI technologies [12]. This study presents a survey conducted among public managers in Mexico and identifies systematic (related to data analysis) and axiological (related to values, ethics, and decisions) capacities from the perspective of public officials [12].