Fighting Financial Crime with Science: Giant Oak CEO Gary Shiffman - A Barefoot Innovation Group Podcast
by Jo Ann Barefoot, on Jun 19, 2019 10:19:22 AM
We know that the money launderers are winning. We know that law enforcement is losing. By now, you’ve probably heard the UN’s statistics on this, that we catch less than 1% of these crimes, which add up to an estimated $1.6 trillion a year. At the heart of this problem is asymmetrical technology. The bad guys have sophisticated, networked, rapidly improving technologies, and the good guys are stuck in the past.
Happily, that is starting to change, and no one is doing more to change it than today’s guest. Giant Oak CEO Gary Shiffman is a behavioral scientist. He and his colleagues are combining behavioral modeling techniques, deep knowledge of illicit activity, and analysis built on AI and machine learning, to try to turn the tide on financial crime. As he puts it, they are democratizing their expertise by converting it into software that everyone can use.
Many patterns of financial crime, including identity authentication issues, can quite easily be found by machines scanning large sets of data, but are hard to detect by humans who are running Google searches to meet AML Know-Your-Customer requirements. These kinds of machine learning techniques are in use every day in other parts of the economy. They are how Amazon or Google guess at what we want to shop for, or Netflix predicts what movie we might like. Now, they are coming into the great fight against financial crime.
In our conversation, Gary describes GOST, which stands for the Giant Oak Search Technology that does large scale screening, vetting, and continuous monitoring of data, and front loads these searches so that risk signals are caught at the start, at the stage of customer screening, rather than after a suspicious situation has arisen and needs investigation. Gary calls this process “reindexing of the internet” -- curating vastly more online information than is captured in a normal search.
He also uses a term I’ve been quoting ever since our conversation -- the “traveling algorithm.” It captures an emerging AML concept we’ve discussed in past shows, which is the idea that instead of bringing the data to the technology in a central place that make it a target for cyberattack, we should take the technology to the data. We can leave it decentralized where it’s safer, and just have everyone share algorithms that can find the big patterns.
When he’s not at Giant Oak, Gary is a professor at Georgetown University here in Washington, D.C., where he teaches a course called, “The Economics of Organized Violence.” Later this year he will have a book coming out on that topic, and you’ll be fascinated to hear him talk about the economics of criminal ventures.
Gary shares my optimism about the regulators’ role in support of AI and machine learning. He points to the interagency statement released by the financial regulators and FinCEN late last year at the ABA Financial Crimes Conference, which proactively encourages the industry to adopt and to test regtech for AML. Since we talked, FinCEN has launched its new innovation initiative and “office hours” program. Like me, Gary believes we need hackathons, experiments, and lots of testing to find out what works, fix weaknesses, and accelerate learning and sharing.
Like me, also, he is confident that this technology will succeed and that we will change the world.
Barefoot Innovation Podcast Episode - Regulation Revolution:The Financial Conduct Authority and Digitally-Native Regulatory Design
Barefoot Innovation Podcast Episode - Regulation Innovation: The FCA’s Christopher Woolard
Barefoot Innovation Podcast Episode - The Future of Regulation: The FCA’s Regtech Leader Nick Cook
More on Dr. Gary Shiffman
Dr. Gary Shiffman is an expert in the economics of national security and is passionate about using social science and cutting-edge network and data analysis to protect the world from illicit actors. His goal is to make it more difficult for criminals and terrorists to accomplish their goals. In order to do this, Dr. Shiffman focuses on understanding institutions and individuals engaged in the non-random production of violence, and then on creating innovative ways to undermine these activities and networks.
In addition to his work as the founder and CEO of Giant Oak, Dr. Shiffman has been a professor at Georgetown University since 2002. His past professional experiences have positioned him as an expert in the unique intersection between the social sciences, big data, business, and national security concerns. These include service as Managing Director of the Chertoff Group, Senior Vice President and General Manager of the Risk Management Solutions business unit at L-3 Communications, and Chief of Staff at U.S. Customs and Border Protection. He has also worked on anti-terrorism and homeland defense issues at an international law firm, advised US Senators as a National Security and Senior Policy Adviser to the US Senate Leadership, and served in policy, planning, and operational positions in the U.S. Department of Defense.
Gary proudly served his country and is a decorated US Navy Gulf War veteran. He earned his Ph.D. in Economics from George Mason University, his MA from Georgetown University in Security Studies and his BA in Psychology from the University of Colorado.