OAK — OnChain Attack Knowledge

Worked example · 2024-01

Blue-chip DeFi yield-protocol laundering via LST and lending-market cover — Ethereum — 2024

Loss
aggregate laundering volume estimated in the tens of millions USD routed through blue-chip DeFi protocols (Aave, Compound, Lido, Maker/Spark) as cover-crowd staging within larger laundering chains during 2024. The pattern was: illicit-proceeds-origin wallet deposits into a major DeFi protocol, holds the position for a period consistent with legitimate yield-seeking behaviour (weeks to months), withdraws to a different address, and routes the withdrawn proceeds to downstream off-ramps — using the protocol's high-volume legitimate-user baseline as statistical cover.
OAK Techniques observed
OAK-T7.006 (DeFi Yield-Strategy Laundering — primary; illicit proceeds deposited into blue-chip DeFi protocols as cover-crowd staging, with the deposit-and-withdraw trajectory designed to mimic legitimate yield-user behaviour) + OAK-T7.003 (Cross-Chain Bridge Laundering — the upstream source-of-funds entered the Ethereum DeFi ecosystem via cross-chain bridges from the source chain of the original extraction) + OAK-T7.001 (Mixer-Routed Hop — in pre-2024 chains, Tornado Cash served the equivalent staging role that T7.006 now serves; the shift from T7.001 to T7.006 as the cover-crowd staging layer is the structural evolution documented here).
Attribution
inferred-strong at the laundering-operator-cluster level. Chainalysis and TRM Labs have attributed specific blue-chip DeFi deposit clusters to known illicit-source wallets (DPRK-attributed thefts, ransomware affiliate wallets, darknet-market vendor wallets). Per-position attribution to named individuals is not asserted.
Key teaching point
The structural evolution from T7.001 (mixer) as cover-crowd staging to T7.006 (blue-chip DeFi) as cover-crowd staging is the defining 2023-2024 trend in sophisticated laundering operations. Tornado Cash deposits were a binary signal of laundering intent (no legitimate user deposits into a mixer); Aave/Compound/Lido deposits are statistically indistinguishable from legitimate yield-seeking behaviour at the per-deposit level. The laundering operator exploits the asymmetry: the cover crowd in a blue-chip DeFi protocol is orders of magnitude larger than a mixer's deposit pool, making per-position behavioural analysis the only viable detection methodology — and per-position analysis requires per-protocol legitimate-user baselines that most protocols do not publish.

Summary

Through 2024, forensic vendors documented a structural shift in sophisticated laundering operations: the cover-crowd staging layer — previously dominated by Tornado Cash (T7.001) — moved to blue-chip DeFi yield protocols. The shift was driven by the post-2022 sanctions environment that made mixer deposits a high-signal detection event, combined with the growing legitimate-user baseline in major DeFi lending markets and LST protocols.

The laundering pattern in blue-chip DeFi protocols followed a consistent structure:

  1. Deposit into a high-volume, high-legitimate-user-baseline protocol. Laundering operators favoured Aave (lending markets), Compound (lending markets), Lido (liquid staking), and Maker/Spark (CDP/lending) — protocols where daily deposit volume is in the hundreds of millions to billions, providing the densest cover crowd. The deposit was typically in a major stablecoin (USDC, DAI) or in ETH.

  2. Position holding with yield-user-mimetic behaviour. Unlike the pre-2024 flash-deposit-then-withdraw pattern, sophisticated operators held positions for weeks to months — durations consistent with the legitimate yield-user distribution. Some operators even claimed yield (calling claimRewards() or equivalent) to strengthen the behavioural mimicry, though the yield-claim rate among laundering positions remained significantly below the legitimate-user baseline.

  3. Withdrawal to a different address than the deposit source. This is the structural T7.006 signal: where the protocol permits a non-self withdrawal recipient, the operator withdraws to a fresh address unlinked to the deposit-source cluster, breaking the on-chain trail within the protocol. For protocols that enforce same-address withdrawals, the operator instead used multi-protocol rotation — depositing in Protocol A, withdrawing, depositing the withdrawn amount in Protocol B — to achieve the same trail-fragmentation effect across protocols rather than within a single protocol.

  4. Downstream routing to off-ramps. The withdrawn proceeds were routed to CEX deposit addresses, OTC desks, or fiat off-ramps, with the "DeFi yield farming" persona providing the source-of-funds narrative.

The cover-crowd density advantage was the load-bearing structural feature. Aave's USDC lending market processes orders of magnitude more daily deposit/withdraw volume than any mixer ever did. A single $500,000 deposit into Aave USDC is submerged in the legitimate deposit flow; the same amount deposited into Tornado Cash was a high-visibility event. The laundering operator leveraged this asymmetry to move from a high-signal cover-crowd staging layer (Tornado Cash, detected at deposit time) to a low-signal cover-crowd staging layer (blue-chip DeFi, detected only through per-position behavioural analysis).

The detection challenge is the per-position behavioural analysis requirement. Distinguishing a laundering position from a legitimate yield-user position requires:

  • Per-position duration distribution analysis (laundering positions cluster at shorter durations).
  • Yield-claim-rate analysis (laundering positions claim yield at below-baseline rates).
  • Deposit-source/withdrawal-recipient divergence analysis (where the protocol permits).
  • Cross-protocol rotation-graph analysis (laundering positions traverse more protocols in sequence).
  • Upstream source-of-funds cluster attribution (laundering positions originate from known illicit-source wallets).

These analyses require per-protocol legitimate-user baselines — deposit-duration distributions, yield-claim rates, withdrawal patterns — that most DeFi protocols do not publish or expose via public APIs. The gap between the detection methodology (well-understood) and the operational data required to execute it (not publicly available at protocol scale) is the binding constraint on T7.006 detection at scale.

Timeline (UTC)

When Event OAK ref
2022-08 Tornado Cash OFAC sanctions; mixer deposits become high-signal detection event; laundering operators begin shift toward DeFi cover-crowd staging (laundering-rail repositioning trigger)
2023 DeFi yield-protocol laundering volume grows; Chainalysis and TRM Labs begin documenting DeFi-as-laundering-substrate at cohort scale T7.006 (emergence of DeFi cover-crowd staging)
2024 Mature T7.006 pattern stabilised: blue-chip DeFi protocols (Aave, Compound, Lido, Maker/Spark) as cover-crowd staging; multi-protocol rotation as trail-fragmentation layer T7.006 (blue-chip DeFi as cover-crowd staging)
2024-Q4 Chainalysis 2024 Money Laundering Report documents DeFi-as-laundering-substrate as a structural trend; per-protocol detection baselines identified as gap (forensic-publication baseline)

Realised extraction

Aggregate laundering volume through blue-chip DeFi protocols estimated in the tens of millions USD across 2024. The volume is distributed across multiple protocols and multiple laundering-operator clusters; no single per-incident loss figure is attributable because the T7.006 layer is a staging-and-cover layer within larger multi-hop laundering chains. The laundering volume represents the nominal value of deposits routed through DeFi protocols as cover-crowd staging; the terminal extraction occurs at downstream off-ramps (CEX fiat off-ramps, OTC desks), not at the DeFi protocol itself.

Public references

  • Cross-reference: T7.006 at techniques/T7.006-defi-yield-strategy-laundering.md.
  • Cross-reference: T7.003 at techniques/T7.003-cross-chain-bridge-laundering.md.
  • Cross-reference: examples/2024-01-post-tornado-defi-yield-laundering.md — post-Tornado-Cash DeFi yield-protocol laundering shift (2024 onward).
  • Cross-reference: examples/2023-2025-defi-yield-strategy-laundering-cohort.md — DeFi yield-strategy laundering cohort (2023-2025).
  • [chainalysis2024laundering] — Chainalysis 2024 Money Laundering Report (DeFi-as-laundering-substrate aggregate cohort framing).
  • [trmlabsdefilaundering2024] — TRM Labs, "DeFi Protocols as Laundering Substrate — 2024 Trends" (2024).

Public References

See citations in corresponding technique file.

Techniques demonstrated (3)