Communications of the ACM
How Technology Can Combat Human Trafficking
Law enforcement organizations across the United States have recently arrested multiple people charged with various crimes that include organizing, operating, or paying for services from human trafficking rings.
To help address the problem, Kristen Abrams, senior director of combatting human trafficking at Arizona State University's McCain Institute for International Leadership, and Nadya Bliss, director of ASU's Global Security Initiative, recently helped organize a conference at the United Nations on how computational science and AI can help combat human trafficking.
Question: Is there an estimate of how many people in the U.S. are trafficked?
Kristen Abrams: Many people would like to know the prevalence of human trafficking in the United States, especially those tasked with combatting this crime. More research and reliable supporting data would be required to arrive at an authoritative estimate. While the U.S. attorney general and the U.S. Department of State compile and release annual reports that contain some statistics regarding U.S. activities to combat human trafficking, they only provide part of the information required to form a full estimate.
Another source that provides some of the information that may help further inquiries into national prevalence is Polaris, a nonprofit organization seeking to eradicate human trafficking, which has operated a national human trafficking hotline for over a decade. In 2017, Polaris reported that it worked on 8,759 cases of human trafficking that involved 10,615 victims of the crime. It should be noted that these are cases and victims that appear to involve human trafficking, but not confirmed victims of human trafficking.
Global estimates help to contextualize U.S. numbers. The International Labour Organization, for example, reported that at any given time in 2016, there were an estimated 24.9 million people trapped in conditions of forced labor globally. The ILO further reports that women and girls are disproportionately affected by human trafficking, representing 99 percent of victims of forced labor in the commercial sex industry and 58 percent in other sectors.
Question: How can existing technology help fight human trafficking?
Nadya Bliss: One huge challenge that was repeatedly raised at the recent Code 8.7 Conference at the U.N. is the lack of a common database—of victims or of traffickers—that can be accessed by law enforcement organizations, governments, and the organizations working to combat trafficking. That's an example of an existing technology that could aid in helping law enforcement detect and follow up on trafficking-related issues, but there are challenges in creating such a database: For example, there would need to be a common data format, and there would need to be agreements on how information can be used to make sure victims' privacy is protected.
Other examples of current technology that can help include systems to track trafficking on the dark web, or to assess the vulnerability of supply chains to forced labor.
Question: How can AI and other new technologies help?
Nayda Bliss: From an AI perspective, I think there's an opportunity to inject tools like pattern recognition and signal detection to identify suspects, potential victims and survivors. But it's not always obvious that someone is being trafficked, so working through things like analysis of massive data sets can take time because signatures of trafficking can be weak.
Another area that is particularly interesting within computer science is the balance of large-scale computation with security and privacy. In different application domains, security and privacy have different significance. In the area of human trafficking, protection of survivors and their confidential information is of paramount importance, and understanding how to build systems with multilevel security access that provide information without sacrificing privacy is incredibly important and a major challenge.
I'd also note it's important to be sensitive when thinking about applying advanced research to any immediate problem. When there is an on-the-ground crisis, it's not necessarily appropriate to inject a new algorithm or test a new system. It is appropriate to think longer-term about what risks are emerging, and from there to think about how the research can be applied to the problem in various places—for example, detection and rescue of survivors or identifying suspects in trafficking chains.
Additional information on trafficking is available from the U.S. Department of Homeland Security's Blue Campaign, the U.S. Department of State's 2018 Trafficking in Persons report, the U.S. attorney general's Trafficking in Persons report, or the ASU McCain Institute for International Leadership's Combatting Human Trafficking website.
No entries found