Data-Driven Justice: Michael Fullilove on Combatting Human Trafficking with AI
World Day Against Trafficking in Persons
This year’s theme for the United Nations World Day Against Trafficking in Persons is “Reach every victim of trafficking, leave no one behind”. According to the International Labor Organization (ILO), there are an estimated 49.6 million victims of human trafficking and modern-day slavery worldwide, and these numbers are growing1. Authorities are struggling with detection and enforcement, as detection rates for human trafficking fell by 11% in 2020, and convictions plummeted by 27%.2 An estimated 10 million more people were victims of human trafficking and modern slavery in 2021 compared to 20161.
Decision intelligence for combatting human trafficking
Decision intelligence is a groundbreaking solution that empowers authorities with advanced analytics to drive data-driven insights and investigative decision-making. Decision intelligence platforms harness advanced data analytics to fuse and analyze massive amounts of data and accelerate investigations.
Interview with Michael Fullilove
In honor of World Day Against Trafficking in Persons, we had the privilege of interviewing Michael Fullilove, an expert on using decision intelligence solutions to combat human trafficking. Michael served 20 years as a Navy SEAL, and has also advised senior government officials on strengthening their capabilities against illicit networks. Michael has also served as the Chief of Operations at DeliverFund, where he played a crucial role in establishing public-private partnerships to empower law enforcement in countering human trafficking networks. He is also the co-author of peer-reviewed papers on leveraging information systems and technology to disrupt human trafficking networks, and is a member of Interpol’s Human Trafficking Experts group.
Read on to hear Michael Fullilove’s unique perspective on combatting human trafficking with decision intelligence, or watch the video below.
How do you use technology and analytics-based solutions to combat human trafficking?
In my work fighting human trafficking, we use a variety of intelligence tools and develop a link analysis model that visually represents the network’s structure, critical nodes, and centers of gravity. This allows for better decision-making regarding the disruption of those networks. We use AI and machine learning tools to help accelerate the process, enabling us to outpace human traffickers and disrupt their networks. While traditional relational databases are useful for trend analysis, they don’t really help the frontline officer. By visualizing the connections between data pieces in a graphical database, we are able to really assist frontline officers and their superiors in understanding the network under investigation and making connections between different pieces of data, leading to more effective strategies for the disruption of human trafficking networks.
How can law enforcement use decision intelligence to take better advantage of data available to them?
When officers go out to conduct interviews and collect data, they must have a way to import and connect the different pieces of investigation data. For instance, if a law enforcement report detailing an incident at a specific location is uploaded into a decision intelligence system, and later on, data emerges from another case that can be linked back to the information in that initial report, it allows for the creation of a network map. This comprehensive mapping helps connect the dots and understand the broader investigative picture.
What tools did law enforcement agencies use before they had AI and decision intelligence platforms?
You know those scenes on TV where law enforcement has all those push pins and strings connecting different pictures? That’s essentially what they used to do in real life, too. They would create files on their computer with different pieces of data, but there wasn’t a good way to connect all of the information.
How have you seen decision intelligence facilitate communication and collaboration between different agencies?
Decision intelligence systems are able to break down the barriers created by jurisdictional boundaries. When officers from different agencies and locations can use the same system and share data, they create connections across geographies, connecting different law enforcement agencies and breaking down silos. It’s like building a network to fight networks, and that’s a game-changer.
In your experience, what is the number one factor that has led law enforcement to adopt decision intelligence?
Demonstrable success. When law enforcement officers are able to see that a particular system increases the speed and effectiveness of their work, and it helps them get more data to prosecutors so that a prosecutor will accept their case, that is what helps drive forward the adoption of new decision intelligence technologies.
What stands in the way of the widespread adoption of decision intelligence among law enforcement agencies?
One key challenge is that law enforcement agencies often have mandated systems they must use. If a decision intelligence platform doesn’t seamlessly integrate into their existing workflow, it can create additional work for already overwhelmed officers, causing law enforcement officers to be less likely to use the decision intelligence system. Another hurdle is lack of awareness and different attitudes towards new technologies, particularly among more old-head officers, many of whom are reluctant to adopt new tools, as they are less familiar with information technology, artificial intelligence, and machine learning. These factors collectively hinder the broader-scale adoption of decision intelligence within law enforcement.
Do you think law enforcement agencies are on their way to widespread adoption of decision intelligence?
Yes, I am seeing movement in that direction. And inevitably, there’s going to be a broader adoption. Technology moves forward. Newer officers are coming in. I mean, most of the younger people of today have grown up with technology. So it’s much easier for them to assimilate it, incorporate it, and adopt it. So as those people move up the ranks, there’s going to be a much greater impetus for adoption of generative AI. There are a lot of applications for that with respect to not only human traffickers for recruiting and grooming, but also for law enforcement in terms of countering human trafficking. And as generative AI continues to develop and mature, I think we’re going to see a lot more adoption of decision intelligence platforms that utilize AI and other advanced technologies.
Do you have any stories that stand out in your experience working with decision intelligence to combat human trafficking?
One incredible story that comes to mind is what we were able to achieve during a decision intelligence training course we provided for a Texas law enforcement agency to aid them in combating human trafficking. Instead of relying on external organizations, our goal was to empower the law enforcement agency with self-sufficiency. As a final training exercise, we conducted an operation, which resulted in the arrest of a trafficker and the rescue of a victim within just eight hours. Thanks to the data gathered during the operation, the prosecutor was able to charge and convict the trafficker within a year, a significantly shorter timeframe compared to the usual lengthy process in human trafficking cases, which can span multiple years. In this case, the timeline was reduced due to effective utilization of data and the use of decision intelligence.
What law enforcement agencies have you worked with?
I’ve collaborated with various US federal and state law enforcement organizations, as well as international agencies, including Interpol, to combat human trafficking. These efforts extend beyond the United States, addressing both local and regional trafficking models, as well as national and international cases. Many times, local human trafficking operations have ties to larger international organizations. For example, the illicit massage industry has ties to organized crime in China and Southeast Asian countries, and many criminal operations in New York have connections to Eastern Europe. I’ve also done work to combat human trafficking on the southern border of the US, involving transnational criminal organizations from Central America.
How does decision intelligence benefit victims of human trafficking?
Currently, prosecutors often require a victim or survivor to testify in human trafficking cases, which can retraumatize them during cross-examination by the defense attorney. This model is flawed because the defense may try to discredit the survivor based on substance abuse, employment issues, or previous criminal history without considering the coercion involved. However, with the wealth of intelligence and data available today, we can build a strong case against traffickers using decision intelligence to analyze intelligence databases and uncover interconnected data points that directly implicate traffickers. By reaching a certain threshold of evidence, we can eliminate the need for victim testimony, saving the victim from the experience of re-traumatization, and resulting in a more objective and data-driven case that is easier for a jury to comprehend.
What future challenges do you foresee for law enforcement with regards to decision intelligence and human trafficking?
The future challenges in law enforcement center around the development of generative AI. Open-source AI architectures make it difficult to determine responsibility for harmful actions caused by AI products. Who is accountable when generative AI creates content that lures and grooms minors? Current policies and laws don’t address this issue clearly. Determining authorship and proving responsibility for AI-generated content will be a significant challenge.
On the positive side, law enforcement can utilize AI to better investigate human trafficking by leveraging AI tools to analyze larger data sets and gain deeper insights into the data, facilitating better decision-making.
Who is winning the race to adopt AI – bad actors or law enforcement?
Law enforcement is engaged in a perpetual cat and mouse game with criminals. Criminals and other bad actors tend to outpace law enforcement in taking advantage of new technology, as they are not subject to the regulations and bureaucratic constraints as law enforcement is. Criminals treat their illicit activities as a business, and are able to focus virtually all of their resources on advancing their operations, while law enforcement agencies have multiple responsibilities. Limited resources further hinder law enforcement’s ability to keep up with bad actors, as funding and procurement processes cause delays.
The good news is that decision intelligence and new information technology platforms are one of the strongest countermeasures that law enforcement has at its disposal, especially due to their facilitation of cross-jurisdictional information sharing.
Learn More about data-driven justice
Decision intelligence and advanced analytics have enormous potential to accelerate human trafficking investigations as well as many other types of investigations. Click here to learn more about NEXYTE, Cognyte’s decision intelligence platform.
Sources: