Privacy and Compliance: Tornado Cash to ZK-KYC
Executive summary
The privacy-versus-compliance debate became the defining regulatory question for crypto between 2022 and 2026, anchored by the OFAC sanctioning of Tornado Cash in August 2022 and culminating in the Fifth Circuit's Van Loon v Treasury ruling in November 2024 that smart contracts cannot be designated as sanctioned property. The post-Van Loon landscape reframed the question: privacy is not legally extinct, but pure-anonymity tools that ignore sanctions screening face existential legal risk. The ecosystem responded with privacy-pool architectures (Aztec Network's Noir, Privacy Pools by 0xbow, Railgun) that offer compliance-aware privacy through ZK-verified non-membership in sanctions sets. The Travel Rule under FATF guidance, MiCA's transfer-of-funds requirements in the EU, and the proliferation of ZK-KYC tools (zPass, Polygon ID, Sismo, World ID) reshaped how privacy and compliance can coexist. By 2026, the consensus is that privacy is not dead but must be programmatic and selectively disclosable, with compliance baked in as a precondition rather than an opposing force.
Tornado Cash, OFAC, and the August 2022 sanctioning
On August 8, 2022, the U.S. Treasury's Office of Foreign Assets Control (OFAC) added Tornado Cash, a non-custodial Ethereum mixer, to the Specially Designated Nationals (SDN) list, citing its use to launder over $7 billion in cryptocurrency including funds from North Korea's Lazarus Group. The action was unprecedented: OFAC had previously sanctioned individuals and entities, but never an immutable, autonomous smart contract. The sanctioning had immediate effects: Circle froze USDC held in Tornado Cash addresses, GitHub took down the Tornado Cash repository, several developers (notably Alexey Pertsev) were arrested in the Netherlands, and U.S. persons faced legal jeopardy for any interaction with the contract. The action raised fundamental questions: can autonomous code be sanctioned? Can a developer be liable for software they wrote that is later misused by sanctioned parties? Are Americans prohibited from interacting with smart contracts they previously deposited into? The crypto industry, civil-liberties organizations (EFF, Coin Center), and several plaintiff groups challenged the sanctioning in court, with the Coinbase-backed suit (Van Loon, et al. v Treasury) becoming the leading case. Pertsev was convicted in May 2024 by a Dutch court but the conviction was partially overturned in 2025, with most of the legal weight shifting to the U.S. cases.
Van Loon v Treasury and the Fifth Circuit ruling
On November 26, 2024, the Fifth Circuit Court of Appeals issued its ruling in Van Loon v Treasury, holding that immutable smart contracts cannot be designated as 'property' under the International Emergency Economic Powers Act (IEEPA), the statute under which OFAC sanctioned Tornado Cash. The court's reasoning emphasized that property under IEEPA requires the capacity to be controlled or owned, and immutable smart contracts (which the Tornado Cash core pools were) lack that capacity. The ruling did not overturn sanctions on individuals (such as Alexey Pertsev or Roman Storm), nor did it bar the government from prosecuting parties who use mixers to launder sanctioned funds. But it established a critical precedent: code itself, when truly autonomous and non-controllable, is not subject to OFAC designation. Treasury declined to appeal to the Supreme Court, instead removing Tornado Cash from the SDN list in March 2025 while emphasizing continued enforcement against bad-actor users. The post-Van Loon landscape is materially different: privacy infrastructure can be built without per-se sanctions risk, but the developers, operators, and users still face KYC/AML and sanctions-screening obligations under existing money-laundering statutes.
Privacy pools and the Aztec/Railgun architecture
The architectural response to the Tornado Cash episode was the privacy-pool concept, formalized in the 2023 paper by Vitalik Buterin, Ameen Soleimani, and others. Privacy pools allow users to deposit assets and prove, in zero-knowledge, that their deposit comes from a 'good' set (non-sanctioned, KYC-verified) without revealing the specific deposit. The mechanism: each user, when withdrawing, provides a ZK proof that their commitment is in a designated 'compliance set' (which excludes sanctioned addresses). This preserves cryptographic privacy while enabling sanctions compliance. Aztec Network, the most prominent privacy-pool project, launched its mainnet in 2024-2025 after extensive R&D on the Noir programming language and the Aztec proof system. By 2026, Aztec supported private DeFi (private DEX, private lending) with compliance-aware privacy at the protocol level. Railgun, a privacy protocol on Ethereum that predates the Tornado Cash sanctioning, restructured around privacy-pool principles in 2023-2024 and grew steadily to multi-hundred-million-dollar TVL with explicit compliance integrations. 0xbow's Privacy Pools deployment offered a more direct successor to Tornado Cash, with explicit screening of deposits via a combined-screening service. The strategic difference: privacy pools accept that compliance is a precondition for privacy, rather than treating them as opposed values.
FATF Travel Rule, MiCA, and the regulatory frame
The Financial Action Task Force (FATF) Travel Rule, which requires Virtual Asset Service Providers (VASPs) to share originator and beneficiary information for transfers above thresholds (commonly $1,000 or $3,000 depending on jurisdiction), became a central compliance burden for crypto exchanges and wallet providers between 2020 and 2026. By 2026, most major exchanges had implemented Travel Rule compliance through industry-standard protocols (TRP, Sumsub Travel Rule, Notabene, Veriscope), with messaging-layer interoperability still imperfect but improving. Self-hosted-wallet transfers remained a friction point: regulators wanted KYC information on the receiving side even when the receiver was a self-custodial wallet, while privacy advocates argued that requirement effectively required everyone to be a customer of a regulated VASP. The EU's MiCA framework, in its Phase 2 implementation through 2025, included transfer-of-funds requirements modeled on the Travel Rule and required self-hosted-wallet identification for transfers above 1,000 EUR. The U.S. equivalent (FinCEN's proposed self-hosted-wallet rule, which never finalized in its original form) faced industry pushback and was significantly modified through 2024-2025. By 2026, the consensus regulatory frame was that VASPs bear primary responsibility for screening and reporting, while pure software (wallets, smart contracts) bear less direct burden as long as good-faith compliance tools (sanctions screening APIs, OFAC list integration) are integrated.
ZK-KYC, decentralized identity, and selective disclosure
The technical complement to privacy pools is ZK-KYC: cryptographic systems that allow users to prove regulatory-relevant attributes (age, jurisdiction, accreditation, sanctions-list non-membership, KYC verification by a regulated provider) without revealing underlying credentials. By 2026, several production systems were in deployment. zPass and similar zk-passport systems leverage the cryptographic attestations in modern e-passports to allow users to prove age and citizenship without revealing identity. Polygon ID, World ID (Tools for Humanity), and Sismo provide broader identity attestation frameworks with privacy-preserving verification. zkLogin (on Sui) and similar systems allow Web2 OAuth credentials to be used to authenticate to onchain identities without revealing the underlying account. The compliance use case is straightforward: a DeFi protocol can require a ZK-KYC attestation (proof of KYC by a regulated provider, proof of non-sanctioned-jurisdiction residence, proof of accredited investor status) before allowing access, without the protocol or other users learning the specific identity. By 2026, several DeFi protocols (notably privacy-aware variants of Uniswap, Aave, and others) integrated optional ZK-KYC for restricted actions. The framework's ultimate promise - that privacy and compliance are complementary, not opposed - rests on whether regulators accept ZK-attestation-based compliance as sufficient. Early signals from MiCA-related guidance and FinCEN advisory opinions through 2025 suggested cautious acceptance.
Recent milestones (2024 to 2026)
The 2024-2026 window was the most consequential stretch for privacy-and-compliance in crypto's history. Pertsev's May 2024 conviction in the Netherlands (later partially overturned); Roman Storm's U.S. trial proceeding through 2024-2025; the Van Loon Fifth Circuit ruling in November 2024; OFAC's Tornado Cash de-listing in March 2025; Aztec Network mainnet maturation; Railgun's continued growth to multi-hundred-million-dollar TVL; 0xbow's Privacy Pools deployment; the EU's MiCA Phase 2 transfer-of-funds requirements going live in 2025; FinCEN's modified self-hosted-wallet rule iterations; the SAB 121 repeal and reinstatement of more constructive U.S. crypto-regulatory engagement post-2024 election; the FIT21 / market structure legislation movement through 2025; the rise of ZK-KYC integrations (zPass, Polygon ID, World ID, Sismo) into DeFi access flows; and the increasing acceptance by regulators of ZK-attestation-based compliance frameworks. By Q1 2026, several major DeFi protocols had piloted or shipped ZK-KYC integrations as a path to U.S. and EU-compliant access. The crypto-native privacy ecosystem (Aztec, Railgun, Privacy Pools, Penumbra on Cosmos, Namada) collectively held over $1B in privacy-preserving TVL, with growth accelerating post-Van Loon.
Risks, ongoing legal uncertainty, and outlook through 2027
Despite Van Loon, ongoing legal risks remain substantial. The Roman Storm trial and any subsequent appeals will further define developer liability for autonomous code. State-level enforcement (the New York DFS, the California DFPI) operates partially independent of federal frameworks and may take divergent positions. International jurisdictions are not bound by U.S. case law, and the Pertsev case in the Netherlands illustrates that European prosecutors may pursue developers under different theories. Sanctions-evasion enforcement post-Van Loon focuses more heavily on user-level prosecution: parties who knowingly use mixers to evade sanctions face direct liability even if the mixer itself is no longer sanctioned. Privacy-pool architectures rely on the integrity of compliance-set generation; bugs or governance attacks on the screening layer could compromise the compliance properties. Looking forward to 2027, the realistic trajectory is that privacy-with-compliance becomes the default architecture for new privacy infrastructure, with pure-anonymity systems remaining a small permissionless niche. ZK-KYC adoption deepens, with regulators increasingly comfortable with attestation-based compliance. The Travel Rule and MiCA-style transfer requirements will continue to shape exchange and wallet implementations. The deeper question - whether private but compliant DeFi can rival public DeFi at scale - will be answered by Aztec, Railgun, and successor architectures over the 2026-2028 window. Early signs suggest yes, but not at the speed maximalists hoped.
Watch points
- Roman Storm trial outcome and subsequent appeals defining developer liability
- Aztec Network and Railgun TVL growth and DeFi protocol integrations
- ZK-KYC adoption in major DeFi protocols and regulator acceptance
- MiCA Phase 2 transfer-of-funds enforcement and FinCEN self-hosted-wallet rulemaking
TL;DR
The Van Loon v Treasury ruling in November 2024 preserved smart-contract privacy from per-se OFAC sanctioning, and privacy-pool architectures (Aztec, Railgun, 0xbow) plus ZK-KYC frameworks (zPass, Polygon ID, World ID) became the consensus path for privacy-with-compliance through 2027.
Get DeFi Intel research in your inbox
Weekly long-form coverage of papers, incidents, jurisdictions, chains, tokens and the people building them. Free tier covers headlines; Pro adds the analyst-grade breakdowns.