The Bank for International Settlements has published a research paper that develops a conceptual framework for assessing how anti-money laundering and combating the financing of terrorism (AML/CFT) regimes perform across payment instruments, from cash and bank deposits to stablecoins and potential retail central bank digital currencies (CBDCs). The paper argues that uneven design and coverage of AML/CFT rules can create regulatory arbitrage opportunities for illicit payments and influence legitimate users’ instrument choice through a privacy–integrity trade-off, and it advocates combining overarching cross-instrument principles with tailored instrument-specific measures. The analysis centres on whether an instrument relies on identifiable intermediaries that can be made responsible for customer due diligence, transaction monitoring and suspicious transaction reporting, and it describes a potential “waterbed effect” in which tightening controls on one instrument shifts illicit activity to another with a lower probability of detection. A European Union case study traces how AML/CFT requirements have expanded across instruments, including an EU-wide cash payment limit of EUR 10,000 under the Anti-Money Laundering Regulation, conditional exemptions from full customer due diligence for certain low-risk e-money products (eg non-reloadable instruments capped at EUR 150 subject to strict use and convertibility constraints), and a broadened crypto perimeter that brings all cryptoasset service providers into scope while prohibiting hosted wallets and other accounts that anonymise customers or obfuscate transactions. For self-hosted cryptoasset wallets, transaction monitoring applies only where a cryptoasset service provider is involved at one end, with providers required to assess and mitigate money laundering and terrorist financing risk for transfers to and from such wallets. The paper also summarises the European Commission’s proposal for a digital euro framework, where online transactions would be intermediated by payment service providers under standard AML/CFT obligations, while offline transactions would follow a bespoke approach based on limited data availability and potential transaction and holding limits set by the European Commission. The paper identifies areas for further research, including comparisons of AML/CFT approaches across jurisdictions, analysis of geographic arbitrage, empirical testing of the “waterbed effect”, and evaluation of technologies that can either enable or help detect illicit payments.