Human trafficking is a form of modern-day slavery that afflicts both advanced and emerging economies. The statistics on this ‘underground’ industry are truly shocking.

With the nightmare beginning for many individuals at an early age, the International Labor Organization estimates that there are 40.3 million victims of human trafficking globally. Over 75% of these are women and girls, and more than 25% are children. They also estimate that forced labor and human trafficking is a 150 billion-dollar industry worldwide, larger than the GDPs of states of Iowa, Kansas, and Hawaii. In the US, more than 40,000 cases of human trafficking have been reported to the Polaris Hotline over the last ten years.

Figure 1: Age and demographics reported by the Polaris Project, based on data received from 10,615 survivors from Jan. 1 2017 to Dec. 21, 2017.

The Internet has made the problem much worse. We’ve all heard vaguely about ‘Silk Road’ and the ‘Dark Web’ where people are allegedly able to buy drugs, guns and even assassinations using cryptocurrencies like Bitcoin. While most of us have never been on the Dark Web (and wouldn’t want to), it Is shocking how much illicit content remains on the Open Web – the web you and I are familiar with and use daily.

Backpage was a classified ad website launched in 2004. In 2011, it was the second largest classified ad listing service on the Internet in the US after Craigslist. Backpage advertised for all kinds of things, including furniture, rent, and…sex, the last through its dating section. Before it was shut down earlier this year, hundreds of millions of ads were posted. The sheer scale of these ads made it impossible to investigate everyone and everything. Instead, it showed just how rampant the problem was, and how the Internet had so conveniently provided an advertising channel for sex rings and providers to exploit the situation.

We’ve all heard technology referred to as a double-edged sword. So, can we use technology to help law enforcement find victims of human trafficking and to gather evidence that would help them prosecute a case? The short answer is yes. It’s already happening.

Over a period of four years, our group at the University of Southern California’s Information Sciences Institute (ISI) developed an Artificial Intelligence system called DIG (Domain-specific Insight Graphs), to make it easy for subject matter experts such as law enforcement to do research. DIG automatically collects data from all over the Web (in this case, sex ads), reads them, extracts key pieces of information (phone numbers, physical attributes, images, addresses), and analyzes all the information to deliver a search engine.

DIG is automatic in that it does not require the subject matter experts to actually know anything about how the technology ‘works’ which means it does not require high-salary, in-demand data scientists to do the job. In other words, it is a good example of ‘democratic’ technology as it is accessible to ordinary stakeholders in society.

DIG, and other systems like it, are now being used by more than 200 law enforcement agencies across the country. The program has already started having an impact. Evidence gathered by law enforcement using these systems have also led to successful sex trafficking prosecutions in the US in recent years.

DIG serves as a good example of technology, particularly Artificial Intelligence, used for social good. In an era when technology is either being thought of as a consumer good or as a necessary evil, efforts like DIG illustrate that technology can be an active participant in ridding our society of social ills. Whether DIG will continue to erode the ecosystem of human trafficking is for time to decide. But the overall effort, of building systems that continue to benefit society, rather than simply generate returns for shareholders, is something that we hope to continue doing in our group.

Dr. Mayank Kejriwal is a researcher at the University of Southern California’s Information Sciences Institute. His research spans multiple areas in Computer Science and Artificial Intelligence, and has been published in over 30 peer-reviewed international venues. Before joining ISI, he received a Masters and PhD from the University of Texas at Austin in 2016, and holds undergraduate degrees in engineering, philosophy, physics, and banking and finance. He is a passionate advocate of using science and technology for social impact and would like to see more stakeholders in society benefiting from technological advances.