Bitcoin mempool scan helps victims of human trafficking
Rebecca Portnoff, a Ph.D. candidate from the University of California, Berkeley, has developed a search engine that will help law enforcement around the world and possibly Interpol to track human traffickers. It can be called an automated open-source intelligence (OSINT) program, because it gathers data open to public and correlates it so that the detective sees the full crime picture, in this case – sex trafficking chain.
The program Ms. Portnoff invented has two methods of identifying pimps: the first one does not involve cryptocurrencies, it is just a correlation of data the criminals expose when placing ads of their services. The main goal here is to find an author of several messages by searching similar patterns in ads. This is called stylometry and looks a bit similar to what the U.S. authorities presumably used to find out who Satoshi Nakamoto is.
The second step, however, is focused on cryptocurrency data. As the details of transactions are open-source (the readers might remember the tracking of WannaCry ransom), the engine designed by Rebecca Portnoff scans all related data and thus finds out the owners of ads placed on various websites and Dark Web marketplaces. The connection of advertisement messages, phone numbers placed in these ads and cryptocurrency transactions gives very good results, the scientists report. Machine learning approach helps to put together messages a human detective might miss, and some traditionalists in the law enforcement might not even think about cryptocurrency transaction tracking.
In a BBC podcast, Rebecca explains that when the Backpage website, her testing ground, had to use cryptocurrency as credit companies rejected to work with it, she realized that the pseudo-anonymous nature of Bitcoin transactions can help identify vast clusters of ads belonging to certain pimps. Bitcoin is based on blockchain, and the blockchain can be viewed as a list of transactions, or a big flow from which one can identify illicit transactions.