Background
Roman Sterlingov, a dual Swedish-Russian national, was accused of creating and operating Bitcoin Fog, a bitcoin “mixer” designed to obscure the relationship between cryptocurrency deposits and withdrawals. The government presented evidence connecting Sterlingov to the service, including internet-protocol analysis, his early interest in bitcoin mixers, Bitcoin Fog’s links to darknet markets, and operational features such as weekly record deletion and randomized fees.
A jury convicted Sterlingov of conspiracy to commit money laundering, substantive money laundering, operating an unlicensed money-transmitting business, and conducting a money-transmission business without a license under D.C. law. The district court sentenced him to 150 months in prison. On appeal, Sterlingov challenged venue, the timeliness and sufficiency of the prosecution, expert evidence, access to proprietary blockchain-tracing source code, several evidentiary and instructional rulings, and the calculation of laundered funds at sentencing.
The Court’s Holding
The D.C. Circuit affirmed the convictions and sentence. It held that venue was proper in the District of Columbia on all four counts because a federal agent conducted undercover Bitcoin Fog transactions from D.C., received funds there, and provided evidence that the service conducted unlicensed money-transmission activity in the District. The transactions, Bitcoin Fog’s continued operation, and its continued profitability also supplied sufficient evidence that the charged conduct and conspiracy continued within the applicable limitations periods.
The court found no abuse of discretion in admitting expert testimony concerning IP-address overlap and Chainalysis Reactor’s clustering of bitcoin addresses. Although Reactor had not been peer reviewed and its error rate was unknown, testimony about its methods, industry acceptance, corroboration, and performance supported the district court’s reliability finding. The court also rejected Sterlingov’s constitutional challenges to the denial of Reactor’s source code, reasoning that machine-generated outputs did not implicate the Confrontation Clause as presented and that Sterlingov had not shown the code was material to his defense.
The panel further upheld the admission of evidence involving the Welcome to Video darknet site and cooperating witnesses, rejected the constructive-amendment claim, and sustained the willful-blindness instruction. At sentencing, the district court permissibly treated approximately $400 million passing through Bitcoin Fog as laundered funds because Sterlingov did not carry his Guidelines burden to separate legitimate funds from criminally derived funds.
Key Takeaways
- Undercover cryptocurrency transfers into and out of the District of Columbia supported venue for money-laundering and unlicensed money-transmission charges.
- Blockchain-tracing expert evidence may satisfy Rule 702 without peer review or a known error rate when other evidence adequately demonstrates the method’s reliability and acceptance.
- Machine-generated analytical outputs are not testimonial statements for Confrontation Clause purposes merely because humans designed the underlying software.
Why It Matters
The decision provides substantial appellate guidance on the prosecution of cryptocurrency mixers, particularly regarding venue, blockchain-clustering evidence, proprietary tracing tools, and the treatment of machine-generated outputs under the Confrontation Clause.
It also underscores the sentencing consequences of failing to identify and substantiate legitimate funds commingled with criminal proceeds. Where separating the two is impracticable and the defendant does not provide sufficient contrary information, the Guidelines permit the court to use the total amount of commingled funds.