TL;DR: Treasury’s GENIUS Act report correctly identifies digital-asset illicit finance as a growing threat, according to Blockaid, but still centres investigation over prevention even as it cites $5.8B in crypto scam losses, $2.8B stolen by DPRK actors, and 3.8 billion monthly transactions. The practical shift is toward pre-transaction interdiction, behavioural detection, and wallet-layer controls before settlement.
Editorial analysis by NHI Mgmt Group, based on content published by Blockaid: “What the U.S. Treasury's Illicit Finance Report Means for Onchain Security”.
By the numbers:
- $2.8B was stolen by DPRK cybercriminals in just 21 months.
- 3.8 billion monthly transactions were recorded as of early 2025, up 96% year-over-year.
Key questions
A: Start by moving risk checks upstream of settlement.
Q: Why do sanctions screens miss many crypto threats that behavioural detection can catch?
A: Sanctions screens depend on known addresses and entities, while modern attackers can generate new wallets and change infrastructure quickly.
Q: What breaks when transaction risk controls sit after settlement?
A: The institution loses the ability to influence the transfer itself and is reduced to evidence collection.
Practitioner guidance
- Implement pre-transaction decision points Place risk evaluation before signing, routing, or settlement in wallets, dApps, and payment workflows so suspicious transfers can be blocked in time.
- Separate screening from behavioural detection Run sanctions and known-entity screening alongside behavioural models that look for wallet sequencing, contract traits, and suspicious interaction patterns.
- Govern wallet and dApp integrations as control surfaces Review external integrations for override rules, audit logging, model inputs, and escalation paths because they now shape enforcement decisions.
Bottom line: Crypto compliance is moving from post-hoc tracing toward prevention at the point of transaction.
Explore further
View Full Forum → | NHI Foundation Course → | Our Services → | Read the full analysis →