AI and Money Laundering

Abstract (From the Foreword)

The key to control of international crime may depend on cutting off the flow of illegal profits to criminal organizations. It is estimated that $300 billion of "dirty money" may be laundered each year, its origin and ownership obscured as it passes through financial institutions and across national boundaries in an effort to hide and protect it from law enforcement authorities. Criminal organizations, like legitimate businesses, enjoy a swift and nearly risk-free conduit for moving money between countries -- wire transfer systems. Illicit wire transfers are easily hidden among the 700,000 mostly legitimate wire transfers that occur daily in the United States, moving well over $2 trillion. OTA was asked by the Permanent Subcommittee on Investigations of the Senate Committee on Governmental Affairs to assess the proposed use of techniques derived from artificial intelligence research to monitor wire transfer traffic and recognize suspicious transfers. Fully automated computer screening of wire transfers was found to be virtually impossible for technical reasons. However, OTA analysts developed and evaluated a number of alternative configurations of technology that, combined with certain legal and institutional innovations, could greatly enhance the capability of law enforcement agencies to detect and prosecute money launders seeking to exploit U.S. financial institutions and wire transfer systems. Although all of these proposed configurations entail some economic and social costs, including possible diminution of financial privacy, strategies are suggested for minimizing these costs while enhancing the potential usefulness of information technology in control of money laundering.

Text

Information Technologies for the Control of Money Laundering is available in PDF and in paper form from the Government Printing Office.

Links

Office of Technology Assessment

---

Feedback Fineprint