How a Carnegie Mellon Grad's AI Tracks Traffickers and Rescues Victims

Compiled by the editorial desk with reference to public statements and event coverage.

At 16, Emily Kennedy's eyes were opened to the grim reality of sex trafficking during a trip to Macedonia, where she saw children cleaning cars at a stoplight, driven by fear of punishment from the Russian mob. That moment planted a seed that would grow into a technology now credited with helping rescue hundreds of trafficking victims across North America.

Kennedy, now 27, shared her journey at the Women in the World summit in New York City, where she detailed the evolution of her software, Traffic Jam. Developed with machine learning experts at Carnegie Mellon University, where she studied humanities and graduated in 2012, the tool trawls online sex ads to spot potential trafficking victims. Many of these ads include photos, which Kennedy's team uses with Amazon Web Services' facial recognition to match against missing person reports or social media images.

The urgency is real: traffickers frequently move victims between cities to avoid detection, making rapid analysis critical. Traffic Jam automates what would be an overwhelming manual task, sifting through thousands of ads to flag those with telltale signs, such as characteristic misspellings.

Since founding Marinus Analytics in 2014, Kennedy has partnered with law enforcement agencies, prosecutors, and select nonprofits, mindful of the software's potential for misuse. The results are tangible: an 88% success rate and the rescue of hundreds of victims in the U.S. and Canada.

Beyond the Headlines: Rethinking Trafficking Data

Traffic Jam's impact extends beyond individual rescues. It's challenging long-held assumptions about trafficking patterns. For instance, Kennedy noted that data from her software debunked the myth that events like the Super Bowl cause a spike in trafficking. Instead, she found that other gatherings, such as CES, the Oracle conference, and the Daytona 500, saw more significant increases in activity.

Kennedy also criticized the reliability of commonly cited trafficking statistics, many of which are outdated and disavowed by their own researchers. By providing fresh, data-driven insights, Traffic Jam is helping to paint a more accurate picture of the problem's scale.

Looking ahead, Marinus Analytics is expanding its reach, with Traffic Jam now crawling data from 18 countries and plans to enter the U.K. market. The company is also exploring applications in other areas of predictive policing, such as illegal gun and drug sales.

Kennedy's work earned her the 2018 Mothers of Invention grant from Toyota, a sponsor of the summit, underscoring the recognition of her innovative approach to a persistent societal challenge.