The alignment of artificial intelligence (AI) systems with societal values and the public interest is a critical challenge in the field of AI ethics and governance. Traditional approaches, such as Reinforcement Learning with Human Feedback (RLHF) and Constitutional AI, often rely on pre-defined high-level ethical principles. This article critiques these conventional alignment frameworks through the philosophical perspectives of pragmatism and public interest theory, arguing against their rigidity and disconnect with practical impacts. It proposes an alternative alignment strategy that reverses the traditional logic, focusing on empirical evidence and the real-world effects of AI systems. By emphasizing practical outcomes and continuous adaptation, this pragmatic approach aims to ensure that AI technologies are developed according to the principles that are derived from the observable impacts produced by technology applications.

The increasing popularity of large language models has not only led to widespread use but has also brought various risks, including the potential for systematically spreading fake news. Consequently, the development of classification systems such as DetectGPT has become vital. These detectors are vulnerable to evasion techniques, as demonstrated in an experimental series: Systematic changes of the generative models’ temperature proofed shallow learning—detectors to be the least reliable (Experiment 1). Fine-tuning the generative model via reinforcement learning circumvented BERT-based—detectors (Experiment 2). Finally, rephrasing led to a >90% evasion of zero-shot—detectors like DetectGPT, although texts stayed highly similar to the original (Experiment 3). A comparison with existing work highlights the better performance of the presented methods. Possible implications for society and further research are discussed.

In recent years, there has been a global trend among governments to provide free and open access to data collected by Earth-observing satellites with the purpose of maximizing the use of this data for a broad array of research and applications. Yet, there are still significant challenges facing non-remote sensing specialists who wish to make use of satellite data. This commentary explores an illustrative case study to provide concrete examples of these challenges and barriers. We then discuss how the specific challenges faced within the case study illuminate some of the broader issues in data accessibility and utility that could be addressed by policymakers that aim to improve the reach of their data, increase the range of research and applications that it enables, and improve equity in data access and use.

This article proposes Bayesian adaptive trials (BATs) as both an efficient method to conduct trials and a unifying framework for the evaluation of social policy interventions, addressing the limitations inherent in traditional methods, such as randomized controlled trials. Recognizing the crucial need for evidence-based approaches in public policy, the proposed approach aims to lower barriers to the adoption of evidence-based methods and to align evaluation processes more closely with the dynamic nature of policy cycles. BATs, grounded in decision theory, offer a dynamic, “learning as we go” approach, enabling the integration of diverse information types and facilitating a continuous, iterative process of policy evaluation. BATs’ adaptive nature is particularly advantageous in policy settings, allowing for more timely and context-sensitive decisions. Moreover, BATs’ ability to value potential future information sources positions it as an optimal strategy for sequential data acquisition during policy implementation. While acknowledging the assumptions and models intrinsic to BATs, such as prior distributions and likelihood functions, this article argues that these are advantageous for decision-makers in social policy, effectively merging the best features of various methodologies.

This paper demonstrates how learning the structure of a Bayesian network, often used to predict and represent causal pathways, can be used to inform policy decision-making.

We show that Bayesian networks are a rigorous and interpretable representation of interconnected factors that affect the complex environment in which policy decisions are made. Furthermore, Bayesian structure learning differentiates between proximal or immediate factors and upstream or root causes, offering a comprehensive set of potential causal pathways leading to specific outcomes.

We show how these causal pathways can provide critical insights into the impact of a policy intervention on an outcome. Central to our approach is the integration of causal discovery within a Bayesian framework, which considers the relative likelihood of possible causal pathways rather than only the most probable pathway.

We argue this is an essential part of causal discovery in policy making because the complexity of the decision landscape inevitably means that there are many near equally probable causal pathways. While this methodology is broadly applicable across various policy domains, we demonstrate its value within the context of educational policy in Australia. Here, we identify pathways influencing educational outcomes, such as student attendance, and examine the effects of social disadvantage on these pathways. We demonstrate the methodology’s performance using synthetic data and its usefulness by applying it to real-world data. Our findings in the real example highlight the usefulness of Bayesian networks as a policy decision tool and show how data science techniques can be used for practical policy development.

Data for Policy (dataforpolicy.org), a trans-disciplinary community of research and practice, has emerged around the application and evaluation of data technologies and analytics for policy and governance. Research in this area has involved cross-sector collaborations, but the areas of emphasis have previously been unclear. Within the Data for Policy framework of six focus areas, this report offers a landscape review of Focus Area 2: Technologies and Analytics. Taking stock of recent advancements and challenges can help shape research priorities for this community. We highlight four commonly used technologies for prediction and inference that leverage datasets from the digital environment: machine learning (ML) and artificial intelligence systems, the internet-of-things, digital twins, and distributed ledger systems. We review innovations in research evaluation and discuss future directions for policy decision-making.

The various global refugee and migration events of the last few years underscore the need for advancing anticipatory strategies in migration policy. The struggle to manage large inflows (or outflows) highlights the demand for proactive measures based on a sense of the future. Anticipatory methods, ranging from predictive models to foresight techniques, emerge as valuable tools for policymakers. These methods, now bolstered by advancements in technology and leveraging nontraditional data sources, can offer a pathway to develop more precise, responsive, and forward-thinking policies.

This paper seeks to map out the rapidly evolving domain of anticipatory methods in the realm of migration policy, capturing the trend toward integrating quantitative and qualitative methodologies and harnessing novel tools and data. It introduces a new taxonomy designed to organize these methods into three core categories: Experience-based, Exploration-based, and Expertise-based. This classification aims to guide policymakers in selecting the most suitable methods for specific contexts or questions, thereby enhancing migration policies.

We apply moral foundations theory (MFT) to explore how the public conceptualizes the first eight months of the conflict between Ukraine and the Russian Federation (Russia). Our analysis includes over 1.1 million English tweets related to the conflict over the first 36 weeks. We used linguistic inquiry word count (LIWC) and a moral foundations dictionary to identify tweets’ moral components (care, fairness, loyalty, authority, and sanctity) from the United States, pre- and post-Cold War NATO countries, Ukraine, and Russia. Following an initial spike at the beginning of the conflict, tweet volume declined and stabilized by week 10. The level of moral content varied significantly across the five regions and the five moral components. Tweets from the different regions included significantly different moral foundations to conceptualize the conflict. Across all regions, tweets were dominated by loyalty content, while fairness content was infrequent. Moral content over time was relatively stable, and variations were linked to reported conflict events.

Even though Sub-Saharan Africa (SSA) is lagging in digital technology adoption among the global average, there is substantial progress in terms of Information and Communication Technology (ICT) access and use, where it plays a crucial role in increasing the quality of life in the regions. However, digital gaps still exist within the continents, even though technology adoption across African nations has shown an increase in progress. This paper aims to explore factors that contribute to different adoption rates among three digital technologies in SSA, specifically mobile phones, fixed broadband, and fixed telephones. The methodology utilizes panel regression analysis to examine data sourced from the World Bank, which consists of 48 SSA countries from 2006 to 2022. The findings show a consistent growth in mobile phone subscriptions, different from fixed telephone and broadband internet that shows stagnant progress. Furthermore, infrastructure, and human capital are the most significant factors in addition to other influencing factors. The results of this study provide the African governments with insightful advice on addressing the digital divide and accelerating their digital transformation.

Data for Policy (dataforpolicy.org), a global community, focuses on policy–data interactions by exploring how data can be used for policy in an ethical, responsible, and efficient manner. Within its journal, six focus areas, including Data for Policy Area 1: Digital & Data-driven Transformations in Governance, were established to delineate the evolving research landscape from the Data for Policy Conference series. This review addresses the absence of a formal conceptualization of digital and data-driven transformations in governance within this focus area. The paper achieves this by providing a working definition, mapping current research trends, and proposing a future research agenda centered on three core transformations: (1) public participation and collective intelligence; (2) relationships and organizations; and (3) open data and government. The paper outlines research questions and connects these transformations to related areas such as artificial intelligence (AI), sustainable smart cities, digital divide, data governance, co-production, and service quality. This contribution forms the foundational development of a research agenda for academics and practitioners engaged in or impacted by digital and data-driven transformations in policy and governance.

Recent developments in national health data platforms have the potential to significantly advance medical research, improve public health outcomes, and foster public trust in data governance. Across Europe, initiatives such as the NHS Research Secure Data Environment in England and the Data Room for Health-Related Research in Switzerland are underway, reflecting examples analogous to the European Health Data Space in two non-EU nations. Policy discussions in England and Switzerland emphasize building public trust to foster participation and ensure the success of these platforms. Central to building public trust is investing efforts into developing and implementing public involvement activities. In this commentary, we refer to three national research programs, namely the UK Biobank, Genomics England, and the Swiss Health Study, which implemented effective public involvement activities and achieved high participation rates. The public involvement activities used within these programs are presented following on established guiding principles for fostering public trust in health data research. Under this lens, we provide actionable policy recommendations to inform the development of trust-building public involvement activities for national health data platforms.

What drives changes in the thematic focus of state-linked manipulated media? We study this question in relation to a long-running Iranian state-linked manipulated media campaign that was uncovered by Twitter in 2021. Using a variety of machine learning methods, we uncover and analyze how this manipulation campaign’s topical themes changed in relation to rising Covid-19 cases in Iran. By using the topics of the tweets in a novel way, we find that increases in domestic Covid-19 cases engendered a shift in Iran’s manipulated media focus away from Covid-19 themes and toward international finance- and investment-focused themes. These findings underscore (i) the potential for state-linked manipulated media campaigns to be used for diversionary purposes and (ii) the promise of machine learning methods for detecting such behaviors.

The REDATAM (retrieval of data for small areas by microcomputer) statistical package and format, developed by ECLAC, has been a critical tool for disseminating census data across Latin America since the 1990s. However, significant limitations persist, including its proprietary nature, lack of documentation, and restricted flexibility for advanced data analysis. These challenges hinder the transformation of raw census data into actionable information for policymakers, researchers, and advocacy groups. To address these issues, we developed Open REDATAM, an open-source and multiplatform tool that converts REDATAM data into widely supported CSV files and native R and Python data structures. By providing integration with R and Python, Open REDATAM empowers users to work with the tools they already know and perform data analyses without leaving their R or Python window. Our work emphasizes the need for a REDATAM official format specification to further enable informed policy debates that can improve policy processes’ implementation and feedback.

As data becomes a key component of urban governance, the night-time economy is still barely visible in datasets or in policies to improve urban life. In the last 20 years, over 50 cities worldwide appointed night mayors and governance mechanisms to tackle conflicts, foster innovation, and help the night-time economy sector grow. However, the intersection of data, digital rights, and 24-hour cities still needs more studies, examples, and policies. Here, the key argument is that the increasing importance of the urban night in academia and local governments claims for much-needed responsible data practices to support and protect nightlife ecosystems. By understanding these ecosystems and addressing data invisibilities, it is possible to develop a robust framework anchored in safeguarding human rights in the digital space and create comprehensive policies to help such ecosystems thrive. Night-time governance matters for the data policy community for three reasons. First, it brings together issues covered in different disciplines by various stakeholders. We need to build bridges between sectors to avoid siloed views of urban data governance. Second, thinking about data in cities also means considering the social, economic, and cultural impact of datafication and artificial intelligence on a 24-hour cycle. Creating a digital rights framework for the night means putting into practice principles of justice, ethics, and responsibility. Third, as Night Studies is an emerging field of research, policy and advocacy, there is an opportunity to help shape how, why, and when data about the night is collected and made available to society.

The complex socioeconomic landscape of conflict zones demands innovative approaches to assess and predict vulnerabilities for crafting and implementing effective policies by the United Nations (UN) institutions. This article presents a groundbreaking Augmented Intelligence-driven Prediction Model developed to forecast multidimensional vulnerability levels (MVLs) across Afghanistan. Leveraging a symbiotic fusion of human expertise and machine capabilities (e.g., artificial intelligence), the model demonstrates a predictive accuracy ranging between 70% and 80%. This research not only contributes to enhancing the UN Early Warning (EW) Mechanisms but also underscores the potential of augmented intelligence in addressing intricate challenges in conflict-ridden regions. This article outlines the use of augmented intelligence methodology applied to a use case to predict MVLs in Afghanistan. It discusses the key findings of the pilot project, and further proposes a holistic platform to enhance policy decisions through augmented intelligence, including an EW mechanism to significantly improve EW processes, thereby supporting decision-makers in formulating effective policies and fostering sustainable development within the UN.

Artificial intelligence (AI) requires new ways of evaluating national technology use and strategy for African nations. We conduct a survey of existing “readiness” assessments both for general digital adoption and AI policy in particular. We conclude that existing global readiness assessments do not fully capture African states’ progress in AI readiness and lay the groundwork for how assessments can be better used for the African context. We consider the extent to which these indicators map to the African context and what these indicators miss in capturing African states’ on-the-ground work in meeting AI capability. Through case studies of four African nations of diverse geographic and economic dimensions, we identify nuances missed by global assessments and offer high-level policy considerations for how states can best improve their AI readiness standards and prepare their societies to capture the benefits of AI.

Sexual and gender–based violence (SGBV) is a multifaceted, endemic, and nefarious phenomenon that remains poorly measured and understood, despite greater global awareness of the issue. While efforts to improve data collection methods have increased–including the implementation of the Demographic and Health Survey (DHS) in some countries–the lack of reliable SGBV data remains a significant challenge to developing targeted policy interventions and advocacy initiatives. Using a recent mixed–methods research project conducted by the authors in Sierra Leone as a case study, this paper discusses the current status of SGBV data, challenges faced, and potential research a pproaches.

Climate change exacerbates existing risks and vulnerabilities for people globally, and migration is a longstanding adaptation response to climate risk. The mechanisms through which climate change shapes human mobility are complex, however, and gaps in data and knowledge persist. In response to these gaps, the United Nations Development Programme’s (UNDP) Predictive Analytics, Human Mobility, and Urbanization Project employed a hybrid approach that combined predictive analytics with participatory foresight to explore climate change-related mobility in Pakistan and Viet Nam from 2020 to 2050. Focusing on Karachi and Ho Chi Minh City, the project estimated temporal and spatial mobility patterns under different climate change scenarios and evaluated the impact of such in-migration across key social, political, economic, and environmental domains. Findings indicate that net migration into these cities could significantly increase under extreme climate scenarios, highlighting both the complex spatial patterns of population change and the potential for anticipatory policies to mitigate these impacts. While extensive research exists on foresight methods and theory, process reflections are underrepresented. The innovative approach employed within this project offers valuable insights on foresight exercise design choices and their implications for effective stakeholder engagement, as well as the applicability and transferability of insights in support of policymaking. Beyond substantive findings, this paper offers a critical reflection on the methodological alignment of data-driven and participatory foresight with the aim of anticipatory policy ideation, seeking to contribute to the enhanced effectiveness of foresight practices.

This study examines Nigeria’s National Information Technology Development Agency Code of Practice for Interactive Computer Service Platforms as one of Africa’s first push towards digital and social media co-regulation. Already established as a regulatory practice in Europe, co-regulation emphasises the need to impose duties of care on platforms and hold them, instead of users, accountable for safe online experiences. It is markedly different from the prior (and existing) regulatory paradigm in Nigeria, which is based on direct user regulation. By analysing the Code of Practice, therefore, this study considers what Nigeria’s radical turn towards co-regulation means for digital policy and social media regulation in relation to standards, information-gathering, and enforcement. It further sheds light on what co-regulation entails for digital regulatory practice in the wider African context, particularly in terms of the balance of power realities between Global North platforms and Global South countries.

In this paper, we provide a systematic review of existing artificial intelligence (AI) regulations in Europe, the United States, and Canada. We build on the qualitative analysis of 129 AI regulations (enacted and not enacted) to identify patterns in regulatory strategies and in AI transparency requirements. Based on the analysis of this sample, we suggest that there are three main regulatory strategies for AI: AI-focused overhauls of existing regulation, the introduction of novel AI regulation, and the omnibus approach. We argue that although these types emerge as distinct strategies, their boundaries are porous as the AI regulation landscape is rapidly evolving. We find that across our sample, AI transparency is effectively treated as a central mechanism for meaningful mitigation of potential AI harms. We therefore focus on AI transparency mandates in our analysis and identify six AI transparency patterns: human in the loop, assessments, audits, disclosures, inventories, and red teaming. We contend that this qualitative analysis of AI regulations and AI transparency patterns provides a much needed bridge between the policy discourse on AI, which is all too often bound up in very detailed legal discussions and applied sociotechnical research on AI fairness, accountability, and transparency.