DARPA To Launch Pre-Crime AML Program

• DARPA has announced plans to launch a pre-crime AML program that aims to deter money laundering before it happens
• Citing North Korea's use of cryptocurrency, the program will likely be applied to digital assets
• The program engages controversial predictive policing practices, known to further arbitrary discrimination

The Defense Advanced Research Projects Agency (DARPA) has filed a special notice on the launch of a „Anticipatory and Adaptive Anti-Money Laundering“ Program (A3ML) that "seeks to revolutionize the practice of anti-money laundering" by developing "algorithms to sift through financial transaction graphs for suspicious activity patterns in attempts to anticipate future activities."

"Want to stop our adversaries from laundering money to evade sanctions, buy weapons, and fund drugs that kill Americans? Me too," writes A3ML program manager David Dewhurst on LinkedIn.

"If successful, A3ML would make it prohibitively expensive for our adversaries to transfer illicit value through the global financial system. the technical hypothesis: illicit finance tactics, techniques, and procedures can be algorithmically extracted from diverse data sources and represented in a generic, sharable form," Dewhurst writes.

The goal of the program is to "eliminate global money laundering by replacing the current manual, reactive, and expensive analytic practices with agile, algorithmic methods," the notice states.

Besides a congressional report on the state of AML and a DOJ indictment alleging a money laundering conspiracy between Mexican and Chinese nationals to finance the Sinaloa Cartel, the notice references a CNN report on North Korea's use of cryptocurrencies to fund its missile program, suggesting that A3ML may additionally be applied in the digital asset sector.

Notably, the blockchain surveillance firm Chainalysis recently corrected its 2023 analysis of stolen funds attributed to the Democratic People's Republic of Korea (DPRK) by almost half, from $1B down to $660M. Chainalysis has not responded to a request for comment whether it had informed the Treasury Department of the correction.

The US Government's AML programs have recently faced criticisms from within the Government's own ranks, with former FinCEN leadership questioning whether reporting frameworks as codified under the Bank Secrecy Act are useful at all.

Dubbed "the world's least effective policy experiment," research suggests that the global effectiveness rate of current AML frameworks lies at around 0.02%, while spending 507x as much money on compliance programs than we are able to recover through them.

As DARPA's notice on A3ML states, "the United States [...] still faces challenges in countering money laundering effectively," despite its already draconian efforts which place all US Americans under blanket suspicion.

With even more surveillance, and the further erosion of the right to the presumption of innocence, we can make AML work, DARPA's logic seems to be.

Current AML programs already disproportionately disfavor marginalized communities, and a pre-crime approach as envisioned by DARPA only sets out to make matters worse.

Pre-crime programs, falling under the category of predictive policing, are trained on a set of already biased data points overwhelmingly punishing those who fall outside of what the Government deems the norm.

As seven members of Congress described the practice in a letter urging US Attorney General Merrick Gauland to halt federal funding for predictive policing programs in 2024, "mounting evidence indicates that predictive policing technologies do not reduce crime. Instead, they worsen the unequal treatment of Americans [...] The continued use of such systems creates a dangerous feedback loop: biased predictions are used to justify disproportionate stops and arrests in minority neighborhoods, which further biases statistics on where crimes are happening."

DARPA has stated that it will announce further steps regarding A3ML by the end of January.

Independent journalism does not finance itself. If you enjoyed this article, please consider donating to our Geyser Fund.