Bitcoin

Artificial intelligence has a long way to go before it resembles anything human, but in the meantime, scientists are already getting A.I. to do a lot for humans right now.

Researchers are using A.I. to track bitcoin being used in illegal sex trafficking.

“The internet has facilitated a lot of methods that traffickers can take advantage of,” Rebecca Portnoff, an A.I. researcher at the University of California, Berkeley, told New Scientist. “They can easily reach big audiences and generate a lot of content without having to reveal themselves.” That’s why Portnoff and her colleagues developed a new tool which uses machine learning to identify payment patterns in illicit ads on Backpage.com — a site commonly used to host online ads for sex work.

The new system follows peculiar or repeated bitcoin transactions that are likely used in sex trafficking, and gives authorities a heads up of which chains of payments could be signs of crimes. Every transaction on Backpage uses bitcoin (credit card companies stopped allowing Backpage to use their services in 2015), and each transaction is publicly logged. Someone just has to trace the transaction to the associated bitcoin wallet and connect that to a pattern of similar ads.

Although Backpage is used to list a lot of different classified ads, it’s also a very big source of sex trafficking. As far back as 2012, the site hosted more than 70 percent online sex ads being listed in the U.S.

“We look at cost of the ad and the timestamp, then connect the ad to a specific person or group,” said Portnoff. “This means the police then have a pretty good candidate for further investigation.”

The tool has the potential for slashing the amount of time and resources investigators must expend in searching for sex traffic crimes and making arrests. In a four-week period test period, the results of which were presented recently at the Conference on Knowledge Discovery and Data Mining in Canada, the new tool scanned 10,000 adverts to identify 90 percent of adverts started by the same author. The A.I. exhibited only a 1 percent false positive identification.

Portnoff and her team are moving forward with several different police forces and organizations dedicated to fighting sex traffic crimes on the internet. If those endeavors prove fruitful, it’s likely the tool will expand to the rest of country, and could be the inspiration for other A.I. tools designed to help identify other kinds of crimes being perpetrated through the internet.