How AI Is Easing Digital Asset Recovery In Fraud Cases
The bipartisan Genius Act, signed into law July 18, paved the way for increasing volume and transparency for digital transactions. This trend, coupled with innovations in AI-fueled asset tracing tools, means that the outlook is brightening for parties to track and recover stolen digital assets.
Even where victims are defrauded of traditional assets, fraudsters often turn to crypto to try and launder their ill-gotten gains. Ironically, while many users of decentralized finance and cryptocurrencies may be motivated by perceived anonymity, as the digital asset infrastructure matures, these very technologies may make it easier for wronged parties to identify and recover their assets.
Goal of Asset Recovery Is to Remedy Losses, But the Practice Faces Systemic Hurdles
In its 2024 report, the Association of Certified Fraud Examiners looked at nearly 2,000 fraud cases globally, tallying $3.1 billion in total losses just from occupational fraud — fraud committed by individuals against their employers. ACFE reckons that occupational fraud drains an average of 5% of revenue every year.[1]
Beyond occupational fraud, U.S. consumers lost $12.5 billion to fraud in 2024, which included $5.7 billion lost to investment scams. These losses represented a 25% increase from the previous year, according to the Federal Trade Commission.[2]
Recouping assets lost to fraud is most often a task of asset tracing and recovery. This is a hybrid investigative and legal process that involves finding the assets and then pursuing civil claims to legally turn them over to their rightful owners. The process can proceed whether or not criminal charges are brought in parallel. The cost/benefit analysis for whether to pursue asset recovery, however, can be constrained by fees for attorneys and investigators, which can soak up much of the recovered amounts.
For cases in the U.S. courts, the so-called American rule dictates that attorney fees are generally not recoverable. That means that even after achieving a favorable judgment, the expenses incurred to trace and recover lost assets must be paid by the plaintiff. And the process of investigating where assets have been secreted, and engaging in litigation to obtain turnover of hidden assets, can be tedious and expensive.
As a result, victims of fraud can be motivated to seek a partial settlement without fully recovering their losses from the wrongdoers. This is a systemic problem because it is a feature of the asset recovery process that makes it harder for victims to get made whole. Moreover, it reflects a perverse incentive whereby if a fraudster can reliably get away with returning an amount less than the losses caused, there is little risk in committing fraud.
As More Financial Transactions Go Digital, New Solutions Arise to Address Old Problems
So-called big data tools like digital forensics, AI and blockchain analysis can now be integrated into asset tracing processes to mine massive amounts of data quickly, identifying irregularities indicative of fraud and filtering out irrelevant observations.
With the help of these tools, attorneys and investigators can uncover financial data more efficiently, streamlining the labor-intensive process of manual data analysis. By saving resources in the process of identifying suspicious transactions, these technologies may make it economically feasible to pursue fraud claims for amounts previously too costly to recover.
These tools will gain relevance as the global share of transactions conducted in digital assets increases. Some legislative developments may help pave the way, for example, where they help to strengthen the sense of security and legitimacy around using cryptocurrency.
The House's bipartisan passage of the Genius Act in a 308-122 vote establishes a framework for stablecoins, i.e., digital currencies pegged 1:1 to the dollar, requiring companies that issue stablecoins to register with federal authority and back their stablecoins with reserves held 100% in liquid assets like cash and U.S. Treasury securities.
Cryptocurrency advocates have praised the bill, as the regulatory clarity it brings is likely to attract investment from banks and companies that have so far avoided cryptocurrency markets because of reckless operations of stablecoins. The passage of the Genius Act amplifies the need for the asset tracing technologies explored in this article, as digital currencies become further embedded in market transactions.
The Technology, for the Nontechnical
To understand how these technologies deliver results, it is worth examining the specific techniques deployed to enhance the efficiency of asset tracing in fraud investigations. Among these, anomaly detection and predictive analysis have emerged as powerful big data tools for identifying suspicious transactions and promise to become more deeply embedded in asset tracing cases as machine learning and artificial intelligence advance.
Anomaly detection, also known as outlier detection, refers to the process of identifying unusual patterns or behaviors that differ from normal activity, flagging outliers from large datasets. Predictive analysis works differently, with organizations gathering historical data and training models on past fraudulent transactions in order to identify similar behavior patterns in the present. Both technologies have the same effect.
Rather than having to scour vast datasets manually, experts can closely scrutinize these flagged transactions, spending more time extracting details of the fraud rather than laboriously attempting to locate it. A 2024 U.S. Department of the Treasury press release reported that its use of machine learning AI resulted in $1 billion in asset recovery from fraud in fiscal year 2024 and pledged to proactively target financial fraud by "leveraging data and emerging technologies" going forward.[3]
This concrete success underscores the reality that AI is no longer a speculative add-on, but a real tool that can and will be applied to future asset tracing cases.
Building on these advances, the growing capability of sophisticated digital tools was dramatically illustrated in June, when the U.S. Department of Justice filed the largest cryptocurrency civil forfeiture complaint to date. Cryptocurrency has gained and continues to enjoy a reputation in criminal circles as untraceable, making it a favored instrument in many fraud schemes.
However, that perception is increasingly at odds with reality. In the complaint filed against $225 million in assets taken from more than 400 victims, law enforcement officers from the FBI and Secret Service acknowledged that to uncover the funds, they used blockchain analysis via "free open source blockchain explorers, as well as commercial tools and services."[4]
Blockchain analysis can be used to uncover the specific virtual addresses behind accounts sending and receiving currency, mapping a picture of these interactions even through complex webs of different accounts belonging to various actors. In the DOJ's June action, these tools allowed investigators to follow stolen funds even through layers of transactions with no "legitimate purpose and merely exist[ing] to complicate tracing efforts."[5]
As AI-based big data tools are being used to find fraud in conventional financial records, investigators can apply blockchain technology to map illicit transactions across decentralized platforms. Fraud in the age of digital transactions has grown increasingly intricate with perpetrators employing elaborate schemes and layers to conceal their activity.
However, innovations that facilitate fraud have been met by technologies that equip investigators and litigants with the means to unravel it and do so in an efficient manner. The use of blockchain tools in this case did not depend upon a limitless investigative budget. At least in part, it leveraged free, publicly available software.
Although it will take some time yet for plaintiffs to realize cost savings from applying big data technology to asset tracing, the pieces are in place for a significant reduction in overall costs through lower attorney and investigative fees — and, hopefully, better results. Digital tools are already delivering gains in productivity.
As this technology develops and becomes more ubiquitous, parties will be able to uncover useful information regarding whether to bring a suit and overall gain a clearer, more specific understanding of financial fraud earlier in the process.
AI tools can be leveraged to identify and present the details of a fraud case before significant expenses are incurred in litigation, which could lead to fewer cases being settled prematurely for the purpose of avoiding fees. In addition, smaller fraud cases could become economically feasible where previously it was too costly to identify fraudulent transactions.
The practical constraints on recouping assets lost to fraud, born of the high cost of tracing and recovering assets, may be set to ease through these new tools.
Artificial intelligence and big data tools have the potential to lower barriers of entry that have historically deterred plaintiffs, empowering a broader range of fraud victims to pursue justice, even in cases that once may have seemed beyond reach.
Solomon B. Shinerock is a partner at Lewis Baach Kaufmann Middlemiss PLLC. He previously served as assistant U.S. attorney in the Northern District of New York and as assistant district attorney at the Manhattan District Attorney's Major Economic Crimes Bureau.
Lewis Baach intern Oscar Cassidy contributed to this article.
The opinions expressed are those of the author(s) and do not necessarily reflect the views of their employer, its clients, or Portfolio Media Inc., or any of its or their respective affiliates. This article is for general information purposes and is not intended to be and should not be taken as legal advice.
[1] https://www.acfe.com/-/media/files/acfe/pdfs/rttn/2024/2024-report-to-the-nations.pdf.
[2] https://www.ftc.gov/news-events/news/press-releases/2025/03/new-ftc-data-show-big-jump-reported-losses-fraud-125-billion-2024.
[3] Treasury Announces Enhanced Fraud Detection Processes, Including Machine Learning AI, Prevented and Recovered Over $4 Billion in Fiscal Year 2024 | U.S. Department of the Treasury.
[4] Largest Ever Seizure of Funds Related to Crypto Confidence Scams | United States Secret Service.
[5] Id.
Click here to read the article on Law360.com (subscription may be required).