OIM - Migration Policy Practice (Volume XV, Number 1)

CONTENTS Connecting migration research, policy and practice worldwide ISSN 2223‑5248 Vol. XV, Number 1, April 2026

MIGRATION

POLICY

PRACTICE 2 Foreword 3 Introduction: Migration governance at a critical crossroads: Insights for the Second International Migration Review Forum Pablo Rojas Coppari, Jenna Nassiri and Micaela Lincango 5 Artificial intelligence as a new frontier in migration governance: The missing piece in the Global Compact for Migration Alina Narusova-Schmitz 12 Hold the line: Countering backsliding on gender ahead of the 2026 International Migration Review Forum and beyond Jenna Hennebry, Varinia Salazar-Barrientos, Paola Cyment and Carolina Gottardo 17 Gender aspects in migration and return: Women’s experiences in Kyrgyzstan and Tajikistan Caterina Francesca Guidi, Vivianne van der Vorst and Ivona ZakoskaTodorovska 23 Innovative pathways to migrant regularization: Lessons from the Opportunity Residence Permit of Germany Zeynep Yanaşmayan and Pau Palop-García 30 IOM and private sector solutions for human rights due diligence to support the World Economic Forum’s Global Data Partnership against Forced Labour Hélène Syed Zwick, Maximilian Pottler and Camila Kirtzman (IOM Thailand) 36 Regularizing Venezuelans in the Andes: Lessons in design and implementation Alessandra Enrico 43 Pathway promises and policy realities: Lessons from the International Student Reset of Canada Lisa Ruth Brunner and Dhriti Mehta 51 Streamlining migrant integration in Peru: The one-stop shop model of public–private Migracentros (2023–2026) Nancy Helena Arellano Suárez and Luisa Feline Freier 59 Beyond aid: The enduring role of diasporas across the humanitarian–development –peace nexus Fabrizia Camplone and Anna ValtolinaVol. XV, Number 1, April 2026

Nancy Helena Arellano Suárez and Luisa Feline Freier 59 Beyond aid: The enduring role of diasporas across the humanitarian–development –peace nexus Fabrizia Camplone and Anna ValtolinaVol. XV, Number 1, April 2026

MIGRATION POLICY PRACTICE

2 In unprecedented times, evidence and data show what makes migration governance effective. While proven tools exist –  including targeted protection mechanisms, labour mobility frameworks and coordinated platforms – political will and sustained investment remain critical gaps. As collective responses become both more necessary and more challenging, sustaining spaces for informed dialogue and evidence-based policymaking is essential. It is in this context, and in the lead-up to the International Migration Review Forum (IMRF) – the principal intergovernmental platform for reviewing progress on the Global Compact for Safe, Orderly and Regular Migration  –  that I am pleased to introduce this special issue of the Migration Policy Practice (MPP). As the international community prepares to take stock of achievements, identify persistent gaps and define priorities for the years ahead, the contributions brought together in this issue offer timely and policy-relevant insights to support these discussions. Foreword The continuation of the journal reflects the enduring commitment of IOM to bridging research, policy and practice, as well as to fostering inclusive, globally representative knowledge that informs decision-making at all levels. By amplifying diverse perspectives and grounded analysis, MPP contributes to strengthening the evidence base needed to navigate an increasingly complex migration landscape, and reinforces the value of multilateral, whole-of-society approaches. I would like to extend my appreciation to the new MPP editorial team for their leadership and dedication, as well as to the Editorial Committee and all contributors whose work enriches this issue. Their efforts exemplify the collaborative spirit that underpins effective migration governance. I trust that this special issue will serve as a valuable resource for those engaging in

new MPP editorial team for their leadership and dedication, as well as to the Editorial Committee and all contributors whose work enriches this issue. Their efforts exemplify the collaborative spirit that underpins effective migration governance. I trust that this special issue will serve as a valuable resource for those engaging in the IMRF and help inform a forward-looking and constructive outcome. Ugochi Daniels Deputy Director General for Operations, IOMVol. XV, Number 1, April 2026

MIGRATION POLICY PRACTICE

3 Migration governance stands at a decisive juncture. Funding cuts to the humanitarian sector, coupled with mounting pressures on multilateralism amid ongoing geopolitical tensions, are straining the broader United Nations system and migrationfocused organizations in particular. In this context, the decision to continue publishing Migration Policy Practice (MPP) reflects the recognition of IOM that supporting migration scholarship can contribute to finding answers to these challenges. Guided by a new Editorial Board and Editorial Committee, MPP remains committed to publishing high-quality research, including from diverse geographies and early career researchers, thereby contributing to dialogue and innovative approaches to policymaking. Through this special issue, we aim to support and inform the second International Migration Review Forum (IMRF), featuring articles aligned with the objectives of the Global Compact for Safe, Orderly and Regular Migration (Global Compact for Migration). The first IMRF, held in May 2022, established the precedent for this State-led, four-year review mechanism mandated by the Global Compact for Migration. It brought together governments, international organizations, civil society, the private sector and migrant communities to evaluate early implementation efforts and chart a course for the Compact’s operationalization. 1 Pablo Rojas Coppari, IOM; Jenna Nassiri, York University; and Micaela Lincango, IOM.

Introduction: Migration governance at a critical crossroads: Insights for the Second International

Migration Review Forum

implementation efforts and chart a course for the Compact’s operationalization. 1 Pablo Rojas Coppari, IOM; Jenna Nassiri, York University; and Micaela Lincango, IOM.

Introduction: Migration governance at a critical crossroads: Insights for the Second International

Migration Review Forum Pablo Rojas Coppari, Jenna Nassiri and Micaela Lincango1 Four years on, the migration landscape has further evolved, introducing new dynamics while amplifying persistent challenges. Technological advances, for instance, have enhanced coordination in humanitarian aid delivery, yet they also expose vulnerable populations to new risks that governance frameworks have not yet adequately addressed. Regularization pathways have expanded in some regions, while in others, irregular status remains a persistent barrier to protection and integration. The current political climate poses challenges to advancing and operationalizing gender-responsive migration policies and practices. Key actors, such as diaspora communities, continue to play vital yet underrecognized roles across the humanitarian– development–peace nexus. Meanwhile, private sector engagement in innovative migration governance solutions is now more critical than ever. Together, these interconnected developments highlight the complexity of delivering on the Global Compact for Migration’s guiding vision. Against this backdrop, this issue brings together contributions from practitioners, researchers, international organizations and civil society actors, each offering timely analysis, grounded reflections and actionable recommendations.

Topics include the following: (a) the role of artificial intelligence in migration governance4

Vol. XV, Number 1, April 2026

MIGRATION POLICY PRACTICE

(Narusova-Schmitz) (Objectives 1, 3, 11, 12 and 17); (b) gender-responsive implementation of the Global Compact for Migration (Hennebry et al.) (Objectives 7, 16 and 17); (c) women’s migration and return experiences in Kyrgyzstan and

17); (b) gender-responsive implementation of the Global Compact for Migration (Hennebry et al.) (Objectives 7, 16 and 17); (c) women’s migration and return experiences in Kyrgyzstan and Tajikistan (Guidi et al.) (Objectives 4, 7, 16 and 21); (d) private sector solutions for human rights due diligence (Syed Zwick et al.) (Objectives 6, 17 and 23); (e) innovative pathways to regularization in Germany (Yanaşmayan and Palop -García) (Objectives 5 and 12); (f) the regularization of Venezuelans in the Andes (Enrico) (Objectives 5, 15 and 16); (g) lessons from the international student pathways of Canada (Brunner and Mehta) (Objectives 5 and 18); (h) integrated service models for migrant inclusion in Peru (Arellano Suárez and Freier) (Objectives 4, 15 and 16); and (i) the enduring role of diasporas across the humanitarian–development–peace nexus (Camplone and Valtolina) (Objectives 19, 20 and 23). As the principal intergovernmental platform for reviewing the implementation of the Global Compact for Migration, IMRF offers a pivotal opportunity for the global community to take stock, recalibrate and renew its collective resolve. It provides a structured space for migration actors to assess achievements, identify gaps, exchange promising practices and articulate forward-looking commitments through the Progress Declaration. This issue therefore aims to serve as a resource for policymakers, practitioners and advocates as they prepare for the discussions that will shape the Progress Declaration and define the next chapter of global migration cooperation. We hope you enjoy this issue. Pablo, Jenna and Micaela5 Vol. XV, Number 1, April 2026 MIGRATION POLICY PRACTICE Artificial intelligence as a new frontier in migration

Declaration and define the next chapter of global migration cooperation. We hope you enjoy this issue. Pablo, Jenna and Micaela5 Vol. XV, Number 1, April 2026 MIGRATION POLICY PRACTICE Artificial intelligence as a new frontier in migration governance: The missing piece in the Global Compact for Migration Alina Narusova-Schmitz1 Introduction Artificial intelligence (AI) is no longer peripheral to migration governance. It is increasingly embedded across the migration cycle, including in border and identity systems, visa and asylum processing, labour recruitment and matching, migration data analysis and service delivery. AI is also reshaping the information environment that influences migration policy, expanding communication opportunities while, at the same time, amplifying risks of misand disinformation.2 Policy and regulatory frameworks, however, have struggled to keep pace with AI deployment, creating governance gaps. While some migrationspecific regulations are emerging at national and regional levels, they remain uneven. At the global level, the Global Compact for Safe, Orderly and Regular Migration (the Global Compact) provides a shared vision for comprehensive migration governance, yet it does not reference AI.3 This omission is increasingly significant, as AI is already affecting the implementation, follow-up and review of the Global Compact. 1 Alina Narusova-Schmitz is a migration policy, governance and strategy expert with two decades of experience at IOM, working across global and regional levels, and advancing multilateral cooperation, including support to the Global Compact for Migration negotiations. Her expertise includes policy development across multiple areas, government advisory support, migration governance and capacity development. She is currently an independent consultant specializing in policy analysis, strategic advice, capacity development and stakeholder engagement. 2 Ana Beduschi and Marie McAuliffe, “ Artificial intelligence, migration and mobility: Implications for policy and practice ” in World Migration Report 2022 (M. McAuliffe and A. Triandafyllidou, eds.) (Geneva, IOM, 2021);

2 Ana Beduschi and Marie McAuliffe, “ Artificial intelligence, migration and mobility: Implications for policy and practice ” in World Migration Report 2022 (M. McAuliffe and A. Triandafyllidou, eds.) (Geneva, IOM, 2021); Ana Beduschi, “Responsible artificial intelligence in international migration management”, Migration Policy Practice, XIV(2):44–51 (2025). 3 United Nations General Assembly resolution 73/195 on the Global Compact for Safe, Orderly and Regular Migration (19 December 2018), paras. 6–7. At the same time, AI governance is accelerating globally. National and regional regulations are multiplying, and international processes – including the United Nations General Assembly resolution on safe, secure and trustworthy AI systems and follow-up mechanisms under the Global Digital Compact –  are establishing new coordination structures. However, these efforts remain largely disconnected from migration governance discussions, even though some regulatory frameworks classify certain AI applications in migration, asylum and border management as high risk.4 Bridging this gap is increasingly urgent. The International Migration Review Forum (IMRF) in 2026 offers a timely opportunity to begin structured reflection on how AI is reshaping migration systems, and to ensure that its integration aligns with the agreed vision of safe, orderly and regular migration, while linking migration governance to the emerging global AI architecture. 4 United Nations General Assembly resolution 78/265 on Seizing the opportunities of safe, secure and trustworthy artificial intelligence systems for sustainable development (21 March 2024); United Nations General Assembly resolution 79/325 on the T erms of reference and modalities for the establishment and functioning of the Independent International Scientific Panel on Artificial Intelligence and the Global Dialogue on Artificial Intelligence Governance (26 August 2025). See, for example, the European Union Artificial Intelligence Act’s classification of certain AI systems in migration, asylum and border control management as high-risk and related analyses of high-stakes migration applications (European Union, Regulation

Scientific Panel on Artificial Intelligence and the Global Dialogue on Artificial Intelligence Governance (26 August 2025). See, for example, the European Union Artificial Intelligence Act’s classification of certain AI systems in migration, asylum and border control management as high-risk and related analyses of high-stakes migration applications (European Union, Regulation (EU) 2024/1689 on laying down harmonised rules on artificial intelligence and amending Regulations (EC) No 300/2008, (EU) No 167/2013, (EU) No 168/2013, (EU) 2018/858, (EU) 2018/1139 and (EU) 2019/2144 and Directives 2014/90/EU, (EU) 2016/797 and (EU) 2020/1828 (Artificial Intelligence Act) (13 June 2024)).6 Vol. XV, Number 1, April 2026 MIGRATION POLICY PRACTICE Artificial intelligence in migration contexts AI in migration is best understood not as a single tool but as a suite of machine-based systems that use data to generate outputs –  such as predictions, recommendations or decisions –  with varying levels of autonomy. Key frameworks, including those of the Organisation for Economic Cooperation and Development (OECD) and the European Union, also note that AI systems may adapt after deployment and pursue explicit or implicit objectives.5 Across the migration cycle, AI is increasingly used in operational decision support and 5 OECD, Recommendation of the Council on Artificial Intelligence (OECD/LEGAL/0449) (22 May 2019); Beduschi and McAuliffe, 2021; European Union, 2024; and United Nations, 2024. administrative workflows. It also shapes public perceptions and narratives by enabling largescale content generation, translation, targeting and amplification. These developments can affect social cohesion and the political space for evidence-based policymaking. There is no comprehensive global inventory of how States apply AI across migration systems; existing analyses provide general overviews or

perceptions and narratives by enabling largescale content generation, translation, targeting and amplification. These developments can affect social cohesion and the political space for evidence-based policymaking. There is no comprehensive global inventory of how States apply AI across migration systems; existing analyses provide general overviews or focus on particular regions and functions. Table 1 provides an illustrative overview of AI application across migration systems and the governance issues these raise. The cross-cutting risks and challenges of such AI usage are discussed in the next section. Table 1. Snapshot of artificial intelligence use in migration systems Functional areas (non‑exhaustive) Examples of artificial intelligence application Main reasons for adoption Key risks and sensitivities Migration data, analysis and policy narratives • Forecasting models • Trend analysis • Early warning systems • Automated data dashboards • AI-assisted policy summaries and reporting • Improve preparedness • Support planning and budgeting • Synthesize large data volumes • Inform decision makers • Support public information • Forecasts are inherently uncertain. • Uncertainty may be undercommunicated to decision makers. • Model assumptions may shape policy framing. • Selective indicators can influence political narratives. • This has the potential to distort the migration evidence base and public understanding. Border and identity systems • Biometric matching for identity verification • Anomaly detection in passenger/border flows • Risk scoring to prioritize screening • Manage high volumes • Increase speed and automation • Improve security and fraud detection • Optimize use of resources • Individuals have limited ability to challenge decisions. • Biometric errors can produce disparate impacts. • Interoperability can propagate errors.7 Vol. XV, Number 1, April 2026 MIGRATION POLICY PRACTICE Functional areas (non‑exhaustive) Examples of artificial intelligence application

ability to challenge decisions. • Biometric errors can produce disparate impacts. • Interoperability can propagate errors.7 Vol. XV, Number 1, April 2026 MIGRATION POLICY PRACTICE Functional areas (non‑exhaustive) Examples of artificial intelligence application Main reasons for adoption Key risks and sensitivities Visas/asylum/ returns • Case triage and prioritization • Document consistency checks and fraud detection • Decision-support tools, including predictive analytics to flag absconding risk or prioritize enforcement and case management • Reduce backlog • Improve consistency • Target scarce resources • Respond to political demand for faster decisions • These systems can have a direct impact on an individual’s status, liberty and protection. • There are concerns regarding due process and explainability. • There is a risk of overreliance on automated outputs. Enforcement and compliance monitoring • Risk scoring to prioritize inspections • Overstay detection systems • Analytics to identify irregular status • Tools supporting detention or removal prioritization • Target limited enforcement resources • Increase compliance • Respond to political pressure • There is a risk of discriminatory or disproportionate enforcement. • These can undermine privacy and due process safeguards. Labour recruitment/ matching • Automated candidate screening and skills matching • Credential verification support • Job-candidate matching platforms • Detection of anomalous recruitment patterns • Address labour shortages • Enable faster matching • Reduce administrative burden • Improve compliance monitoring • There is a potential to reproduce labour-market discrimination. • Some ranking processes remain opaque. • These systems involve extensive personal data processing across borders. Migrant

• Enable faster matching • Reduce administrative burden • Improve compliance monitoring • There is a potential to reproduce labour-market discrimination. • Some ranking processes remain opaque. • These systems involve extensive personal data processing across borders.

Migrant services/ integration • Multilingual chatbots for information and referrals • Targeted outreach based on needs profiling • Analytics to optimize service delivery and integration support • Improve accessibility • Reduce transaction costs • Scale service delivery • Support integration outcomes • There are profiling and privacy concerns. • There are uneven digital literacy and access. • There is a risk of exclusion from essential services.

Note: The above examples are indicative and non-exhaustive; adoption, maturity and safeguards vary widely across contexts.

Sources: Table is compiled based on the following sources: Beduschi and McAuliffe, 2021; OECD, 2019; Petra Molnar, “Artificial intelligence in migration management: Opportunities, challenges, and risks ”, Federal Agency for Civic Education (4 September 2025); and Francesca Diaz, Georgia Ross, Tamiliniyaa Rangarajan and Wael Frikha, Leveraging the integration of artificial intelligence (AI) into border management systems: A best practices report (IOM and Graduate Institute of International and Development Studies, 2025).8

Vol. XV, Number 1, April 2026 MIGRATION POLICY PRACTICE Across these and other migration domains, governments cite similar motivations for adopting AI: managing increasing mobility volumes, enhancing security and fraud detection, reducing backlogs and costs, and responding to political demands for faster decisions. Automated tools are also introduced with the expectation that they will standardize decision-making and reduce variation associated with workload pressures or human error. 6 These considerations help explain the rapid integration of AI across multiple migration areas. However, the governance implications differ depending on the extent to which systems influence legal status, liberty and

they will standardize decision-making and reduce variation associated with workload pressures or human error. 6 These considerations help explain the rapid integration of AI across multiple migration areas. However, the governance implications differ depending on the extent to which systems influence legal status, liberty and access to protection. The breadth of AI deployment means that it affects implementation of the Global Compact as a whole, with the most evident implications for objectives related to migration data and evidence (Objective 1); access to accurate information (Objective 3); addressing vulnerabilities in migration (Objective 7); procedures for appropriate screening, assessment and referral (Objective 12); and eliminating discrimination and promoting evidence-based public discourse (Objective 17). As AI becomes embedded in high-impact decisions, transparency, oversight, accountability and rights safeguards become increasingly central. 6 Beduschi and McAuliffe, 2021; Diaz et al., 2025; Petra Molnar, T echnological testing grounds: Migration management experiments and reflections from the ground up (European Digital Rights and Refugee Law Lab, 2020); and Petra Molnar and Lex Gill, Bots at the gate: A human rights analysis of automated decision-making in Canada’s immigration and refugee system (International Human Rights Program and the Citizen Lab, 2018). Artificial intelligence in migration governance: Core risks and governance challenges The integration of AI into migration governance is fraught with the same challenges in other public sectors, including opacity, bias and concerns related to privacy and accountability. However, these challenges are often amplified in migration and asylum contexts, where decisions can have high-impact and potentially irreversible consequences affecting legal status, liberty and access to protection. Individuals involved may also face significant barriers, including limited legal support or capacity to challenge administrative outcomes. For this reason, several regulatory frameworks classify certain AI applications in migration, asylum and border management as high risk. Some of the major AI-related risks and challenges are outlined as follows.7 Opacity and automation bias

also face significant barriers, including limited legal support or capacity to challenge administrative outcomes. For this reason, several regulatory frameworks classify certain AI applications in migration, asylum and border management as high risk. Some of the major AI-related risks and challenges are outlined as follows.7 Opacity and automation bias The black-box problem –  the difficulty of understanding how AI systems produce outputs  –  creates significant due process concerns. When an AI-enabled system flags a case as high-risk or influences prioritization, applicants may be unable to understand why a decision was taken or how to challenge it effectively, raising due process issues relevant to Global Compact commitments. Opacity also presents a risk of policy inertia: systems built on historical data and embedded design choices may not adapt easily to evolving legal standards or policy priorities. 7 Key sources discussing AI-related risks and ways to address them include the following: Beduschi and McAuliffe, 2021; Diaz et al., 2025; Molnar, 2025; Molnar and Gill, 2018; OECD, 2019; Office of the United Nations High Commissioner for Human Rights and University of Essex, Digital border governance: A human rights-based approach (September 2023); and Office of the United Nations High Commissioner for Refugees, UNHCR AI approach (August 2025).9 Vol. XV, Number 1, April 2026 MIGRATION POLICY PRACTICE Human oversight is frequently proposed as a safeguard. Yet research on automation bias shows that officials may defer to algorithmic outputs, particularly under time pressure or heavy caseloads. When this occurs, review may become procedural rather than substantive. As AI is embedded across multiple stages of migration-related decision-making, black-box dynamics and deference to automated outputs can have cumulative effects. Bias and discrimination AI systems reflect the data and assumptions embedded in their design and training. In migration contexts, this creates risks of unequal impacts across groups. Biometric systems,

migration-related decision-making, black-box dynamics and deference to automated outputs can have cumulative effects. Bias and discrimination AI systems reflect the data and assumptions embedded in their design and training. In migration contexts, this creates risks of unequal impacts across groups. Biometric systems, for example, have demonstrated uneven performance across demographic groups, potentially leading to erroneous matches, heightened scrutiny or detention. More broadly, tools trained on historical administrative data may reproduce existing enforcement patterns, even when relying on seemingly neutral indicators. Without robust testing, monitoring and independent oversight, AI can exacerbate unequal treatment and undermine commitments to non-discrimination and rights-based approaches reflected in the Global Compact. Interoperability, cascading error and function creep Migration governance increasingly relies on interconnected databases and cross-border information-sharing. As systems become more integrated, errors can spread across platforms and become difficult to detect or correct. Interoperability also raises the risk of function creep, whereby data collected for one purpose – such as service delivery or registration – are repurposed for enforcement.

Exacerbating factors: Vendor dependence and uneven capacity These governance challenges are compounded by institutional capacity constraints. Migration authorities often rely on external vendors for system design and maintenance and may lack the technical capacity required for effective oversight and auditing. Globally, regulatory capacity varies widely; in lower-capacity settings, AI systems may be adopted without adequate safeguards.

Managing these risks is essential if AI is to support good migration governance. The next section maps the multilayered and fragmented governance landscape, situating migrationspecific approaches alongside the rapidly evolving global AI governance architecture.

State of play: A multilayered and fragmented governance landscape In response to expanding AI use in migration systems, a patchwork of governance approaches is emerging, ranging from binding regulation to soft-law guidance. While these initiatives strengthen accountability, their uneven development risks deepening fragmentation in a cross-border setting.10

fragmented governance landscape In response to expanding AI use in migration systems, a patchwork of governance approaches is emerging, ranging from binding regulation to soft-law guidance. While these initiatives strengthen accountability, their uneven development risks deepening fragmentation in a cross-border setting.10 Vol. XV, Number 1, April 2026 MIGRATION POLICY PRACTICE Emerging artificial intelligence governance in migration contexts Migration-relevant frameworks can be grouped into three approaches.8 First, a risk-based regulatory approach, mostly developed in Europe and parts of North America, applies tiered safeguards based on the potential impact of AI systems. The European Union’s Artificial Intelligence Act is the most advanced example, classifying certain AI uses in migration, asylum and border management as high risk and requiring safeguards such as human oversight, documentation and monitoring, while prohibiting practices deemed incompatible with fundamental rights. Complementary instruments, such as the Council of Europe’s Framework Convention on Artificial Intelligence, reinforce requirements for impact assessment and accountability. Second, an innovation-first model is applied in the Gulf region. It relies primarily on voluntary compliance and ethical self-regulation by developers and State agencies, emphasizing innovation over precautionary restrictions. The Guiding Manual on the Ethics of Artificial Intelligence Use for Gulf Cooperation Council member States illustrates this