Ukraine

 

We received funding from the National Science Foundation D-ISN/RAPID for a project on "Data Collection for Human Trafficking Recruitment and Responses in Forced Migration and Operations Engineering" (Award CMMI-2330311). Our project on Ukraine focused on crises, such as natural disasters and military conflict exacerbate vulnerabilities to, and manifestations of, human trafficking creating conditions ripe for exploitation. The full-scale Russian invasion of Ukraine in 2022 exemplifies how war, displacement, and economic instability amplify inequalities and increase vulnerabilities to human trafficking – both within Ukraine and globally.  Over 8.1 million Ukrainian refugees left the country of which 90% women and children, and millions are displaced internally. Many lack stable housing, social support, language skills and employment, increasing their susceptibility to exploitation and trafficking. The unprecedented scale of the humanitarian crises in Ukraine and the  adaptive responses  necessitated presented a rare opportunity to gather data regarding human behavior and understand shifting response landscapes. 

Thus, the goal of NSF Award was to collect, compile, and analyze time-sensitive data – both online and interviews -  along the entire trafficking spectrum,  from recruitment, transfer, exploitation, to exit.  We employed a multi-method data collection approach, combining qualitative interviews, social media analysis, and hotline data to gain a comprehensive understanding of vulnerabilities, potential recruitment patterns and intervention effectiveness. We collected social media and targeted websites  to identify potential online grooming and recruitment techniques used to exploit vulnerable individuals.. We also surveyed the response landscape by conducting interviews with international and local organizations that operate hotlines, shelters, and support programs for refugees and trafficking survivors. This data collection has led to several papers which are currently in progress or under review.

Machine Learning Solutions in Identifying Potential Human Trafficking Recruitment During Humanitarian Disasters: Case Study of Ukraine (Under Review)

Humanitarian disasters increase vulnerability to human trafficking. Traffickers exploit social media platforms, such as Telegram, with deceptive offers of housing, work, or aid during disasters. Existing detection methods of trafficking activity are ineffective against high-volume, multilingual communications, hampering identification of emerging trends and recruitment tactics to inform prevention and intervention efforts. Previous research focused on static platforms(e.g. Backpage, Yelp) and monolingual datasets, whereas this study addresses dynamic and multilingual communications in social media including non-textual elements such as emojis. Furthermore, we specifically target recruitment tactics during disaster contexts. Our work advances operations management methods by developing a framework that considers multilingual datasets, emoji standardization, and ensemble learning. In practice, this model can be to enable quicker identification of deceptive trends and inform appropriate interventions. We developed a hybrid machine learning framework that integrates classical models and transformer-based models through a max-voting mechanism. In doing so, the framework takes advantage of the complementary strengths of both model types to optimize performance. Our approach processes 482K multilingual Telegram messages, translates non-English content, standardizes emojis, and employs data balancing techniques, including back-translation and rephrasing, to mitigate class imbalance. Our model achieved high Precision and F1 Scores. Our results confirm that the proposed methods are effective in developing models to detect the potential of human trafficking recruitment online during humanitarian disasters. Managerial implications: This framework enables prioritization of preventative interventions and investigations by automating the identification of potential human trafficking in large-scale multilingual data. It reduces dependence on manual classification, accelerates response times during crises, and adapts to evolving recruitment tactics before exploitation occurs. Moreover, it can also be applied to other humanitarian crises.

Human (In)Security: Counter Trafficking and Migration Hotlines During War (Under Review)

Human trafficking hotlines serve a critical function during wars by providing essential security infrastructure and disseminating information on safe migration and where to seek help in destination countries for war-displaced populations offering an alternative platform for anyone to seek assistance. We analyze migration flows and associated vulnerabilities during war through 88,939 calls received by the National Toll-Free Migrant and Anti-Trafficking Hotline 527 from September 2021 to September 2023-before and after the full-scale invasion. Our study explored who accesses the human trafficking hotline during war and how does insecurity for potential migrants manifest during war as they navigate displacement? We determined that hotline data can be a precursor to the human trafficking patterns that follow because it can identify trafficking trends before victims are identified due to the one-two year lag in identification during war. Hotline data also revealed that the oblasts with the largest number of potentially identified victims also had the highest number of calls to the hotline suggesting that potential victims align with frequency of calls from different oblasts.

Chapter 5. Human Trafficking and Forced Criminality in Wartime in an edited volume Crime and Justice in Wartime Ukraine edited by Yuliya Zabyelina, Oleksii Serdyuk, and Anna Markovska and under contract with Routledge.

This chapter examines human trafficking in the context of war and armed conflict, with a focus on Russia’s 2022 full-scale invasion of Ukraine. This war has triggered mass displacement, economic instability, and widespread vulnerability, creating new opportunities for exploitation. Drawing on qualitative interviews with NGOs and large-scale analysis of online platforms used by displaced Ukrainians, the chapter identifies emerging patterns of human trafficking and forced criminality. In occupied territories, interview data reveal instances of civilians coerced into labor by Russian forces, such as mechanics repairing military vehicles or doctors treating wounded soldiers. It shows how victims may also choose to refrain from reporting these abuses due to fears of being labeled collaborators, resulting in underreporting and gaps between law enforcement and victim services.

Online Human Trafficking Recruitment During Russia's War in Ukraine (In Progress)

This article  analyzes the suspicious Telegram data and compares that to interviews from anti-trafficking stakeholders focusing on recruitment dynamics in Ukraine during the full-scale invasion. We examine what recruitment looks like during the war in Ukraine. We found that approximately 15.5% of the total messages on the different Telegram groups were suspicious and could be potential cases of recruitment for human trafficking. The paper discusses the array of suspicious recruitment messages that we found in the Telegram Groups including posts over the first 18 month’s with word frequencies, levels of vulnerability, and posts by potential traffickers and victims. Then we compare this with interview data focused on what 10 different organizations working on the ground in Central and Western Ukraine see with human trafficking recruitment. These organizations are in the border regions of Western Ukraine and formally occupied territories in Central Ukraine and show what recruitment methods and trafficking dynamics their clients have experienced. They saw evidence of forced deportations, illegal conscription, and abductions and while attention has fallen on sex trafficking, forced labor including forced begging is more prevalent. It is difficult to know what is happening in the occupied territories but we found that people in de-occupied territories are reticent to report trafficking by the Russian army because it can lead to arrests for collaboration with the occupying forces.