This article provides an interdisciplinary exploration of the complex dynamics between artificial intelligence (AI) and inequality, drawing upon social sciences and technology studies. It scrutinises the power dynamics that shape the development, deployment, and utilisation of AI technologies, and how these dynamics influence access to and control over AI resources. To do so, we employ Margaret Archer's social realism framework to illuminate the ways in which AI systems can reinforce various forms of inequalities. This theoretical perspective underscores the dynamic interplay between social context, individual agency, and the processes of morphostasis and morphogenesis, offering a nuanced understanding of how inequalities are reproduced and potentially transformed within the AI context. We further discuss the challenges posed by the access and opportunity divide, privacy and surveillance concerns, and the digital divide in the context of AI. We propose co-ownership as a potential solution to economic inequalities induced by AI, suggesting that stakeholders contributing to AI development should have significant claims of ownership. We also advocate for the recognition of AI systems as legal entities, which could provide a mechanism for accountability and compensation in cases of privacy breaches. Finally, we conclude by emphasising the need for robust data governance frameworks, global governance, and a commitment to social justice in navigating the complex landscape of AI and inequality.

Based on 20 countries across Europe, North America and North Africa, this report synthesises key trends and patterns of national policy approaches towards migrant irregularity, highlighting commonalities and differences across various contexts. In particular, this report examines three key research questions: how have irregular migration policies evolved over time and in response to what; what pathways into and out of irregularity have these policies produced or aimed to address; and what challenges have hindered policy implementation. In doing so, the report aims to contextualise irregular migration policy changes, as well as how such policies can channel migrants into or out of irregularity.

This working paper seeks to utilise mortality data, when linked to population register data, to assess the potential of such data in developing robust estimators for hard-to-reach groups, specifically undocumented migrants. Recognising the gaps in current migration statistics, the study proposes this novel approach as a means to offer more accurate and nuanced indicators of irregular migration. Addressing the existing challenges related to international migration data, such as its incompleteness, lack of recency, and the need for harmonization, the paper asserts that mortality data can provide valuable insights into these issues, further shedding light on the demographics, living conditions, and health access of undocumented migrants. Using Belgium as a case study, the research demonstrates how death registry data can be employed to develop migration indicators. By applying this method, we can better identify and understand the characteristics of unregistered populations, offering age-and gender-specific results that could significantly inform future policy decisions. By investigating the mortality paradox among migrants, the paper reveals fascinating trends that offer deeper insights into the factors influencing migrant mortality. The study posits that the mortality extrapolation methodology could serve as a crucial tool to fill in the knowledge gaps in migration statistics, and hence, a valuable source of harmonised statistics on irregular populations in Europe. However, the paper also acknowledges the necessity for improved data quality and the development of more advanced statistical tools to effectively analyse this type of data. Given the piecemeal nature of current migration information, it emphasises the importance of broadening our comprehension of the mortality-migration relationship to enable the formulation of more effective, evidence-based policy decisions.

To gain a better understanding of migration patterns and trends, policymakers, researchers, and analysts require high-quality data on migration, including the number of migrants, their characteristics, and the reasons for migration. This information is crucial for developing effective migration policies and programs, and for monitoring and evaluating their impact. However, there have been significant gaps in international migration data for several decades. National statistical institutions (NSIs) have a crucial role to play in collecting and reporting data on international migration. They are responsible for ensuring the quality and completeness of migration data, which is essential for policymakers. Understanding the challenges faced by NSIs in collecting and reporting migration data can help improve the quality of data and inform policy decisions. However, the NSIs’ perspective is often overlooked in academic research. By introducing the “problem-centered institution questionnaire methodology”, this paper provides a cross-country analysis of the challenges that NSIs face in collecting and reporting international migration data. Drawing insights from 30 countries, the study finds that there are significant gaps in the quality and completeness of migration data, particularly in countries without legal responsibility for data collection. NSIs play a crucial role in improving the quality of migration data, but this requires time and political attention. The study suggests that shifting to administrative sources can help gather better data in a timely fashion. Experience sharing and cooperation across NSIs can also help address gaps in the data. In many countries, stock data are derived from censuses and large-scale surveys.

Migration is one of the key aspects of the Sustainable Development Goals (SDGs). To understand global migration patterns, develop scenarios, design effective policies, focus on the population's needs, and identify how these needs change over time, we need accurate, reliable and timely data. and the United Nations High Commissioner for Refugees (UNHCR). Our results demonstrate that the gaps could be categorized under (1) definitions and measures, (2) drivers or reasons behind migration, (3) geographic coverage, (4) gaps in demographic characteristics and (5) the time lag in the availability of data. The reviewed sources suffer from the gaps, which are not mutually exclusive (they are interlinked): the quality and availability of both migration flows and stocks data vary across regions and countries, and migration statistics highly rely on immigrants' arrival.