Improving Anti-Money Laundering Models with Synthetic Data

As readers of this blog are well aware, an effective anti-money laundering (AML) regime is crucial for fighting grand corruption, as well as other organized criminal activity. A key part of the AML system is the requirement that banks and other financial institutions identify suspicious transactions and file so-called suspicious activity reports (SARs) with the appropriate government agencies. This is an enormous task, given the volume of financial transactions that banks need to monitor and the challenge of identifying which of those transactions ought to be considered suspicious. Banks spend billions on AML compliance every year, and have developed complex automated systems to assist them in flagging suspect transactions, but existing systems’ ability to efficiently sort suspicious from innocent transactions is limited by the sheer complexity of the task. (False positive rates with current systems, for example, frequently top 90%.)

Many believe that artificial intelligence (AI) systems, such as those employing machine learning (ML), hold enormous promise for improving AML compliance and reducing cost. ML algorithms scrutinize vast datasets to identify patterns that can be used to fashion predictive models. In the AML context, ML algorithms identify those transaction characteristics (or complex combinations of transaction characteristics) that are associated with money laundering, and use these patterns to more efficiently and effectively identify suspicious transactions.  

But some commentators have suggested reasons for skepticism, or at least caution. For example, Mayze Teitler recently wrote on this blog about a number of challenges to operationalizing AI-derived algorithms in the AML context, primarily those arising from limitations in the data on which those algorithms are based. As Mayze correctly pointed out, ML algorithms require vast datasets from which to learn, and the data demands are compounded by the relatively rarity of known money laundering cases in the existing datasets.

Despite these concerns, I am more bullish than Mayze regarding the promise of AI-based AML systems. Many of the challenges and concerns regarding the development of effective AI systems in the AML context can be overcome through the use of synthetic data.

Continue reading

To Fix the United States’ Corrupt Border Agency, Defeat Its Union

Immigration reform is likely to be a high priority for the Biden Administration, and while most of the attention will focus on substantive reforms and enforcement strategy, the agenda should also include rooting out corruption in U.S. Customs and Border Protection (CBP), the agency charged with protecting the United States’ land borders. CBP is the nation’s largest federal law enforcement agency. It is also among its most corrupt. Border Patrol agents and CBP officers are regularly arrested—at a much higher rate than other federal law enforcement personnel—for a variety of corrupt activities, including accepting bribes, smuggling drugs, collaborating with organized crime groups, and selling government secrets. (In one case, a Border Patrol agent even gave a cartel member a literal key to a border gate.) All told, U.S. border guards accepted an estimated $15 million in bribes over the 2006–2016 period. Senior CBP officials have estimated that as many as 20% of CBP employees may be corrupt, and almost half of CBP personnel say that they’ve witnessed four or more acts of misconduct by their colleagues in the preceding three years.

The story of CBP’s corruption has been well told, including in voluminous investigative reporting, an advisory panel report, and congressional hearings. Yet little has changed. And this is not because nobody has figured out what policy reforms could make a difference. Indeed, experts who have studied the problem have laid out, clearly and consistently, a package of recommendations that would make a substantial difference. That package includes two main elements. First, CPB must devote more resources to monitoring and investigating CBP personnel. For example, the agency should hire substantially more internal affairs investigators; subject exiting personnel to regular reinvestigations (including periodic polygraph examinations); and equip all officers and agents with body cameras and mandate their consistent use. Second, leadership must reform CBP’s culture, which too often tolerates bad actors and punishes whistleblowers, and must provide better training in how to respond to misconduct.

The failure to address the CBP’s corruption problem, then, has not been due to a lack of viable, feasible reforms. The main problem is political—perhaps most importantly, the entrenched opposition of the National Border Patrol Council (NBPC), the powerful union that represents Border Patrol agents. The NBPC has systematically blocked efforts to crack down on corruption. Indeed, according to James Tomsheck, who led CBP’s internal affairs unit from 2006­–2014, NBPC leadership opposed each and every one of his integrity proposals over his eight year tenure. (For example, the union opposed CBP’s initiative to proactively identify corrupt officers and agents through polygraphing.) If the Biden Administration is serious about rooting out CBP corruption, it will need to take on the NBPC.

Continue reading