Last month, a cross‑chain bridge linked to a protocol called Kelp DAO was exploited and the atackers (presumed to be the North Korean hacking outfit Lazarus) got away with something the region of $300m in cryptocurrency. The hack itself was a useful case study in decentralised finance (DeFi) protocols, but what has happened since iis a useful case study in money.
The Kelp exploit was based on a form of a staked‑Ethereum product known as rsETH, which is a liquid‑staked Ethereum token. When users want to participate in Ethereum’s proof‑of‑stake network, they can either stake their own ETH directly (which locks it up) or deposit ETH into a liquid‑staking protocol that issues a token such as rsETH in return. Each unit of rsETH represents a claim on a corresponding share of staked ETH plus the rewards generated by that staking pool. Now, rsETH is designed to be easy to trade or lend, just like any other token.
So: borrowers use rsETH as collateral because it is pegged to the value of ETH, and lenders are comfortable accepting it on that basis. The exploit mentioned above arose because a flaw in the Kelp‑LayerZero bridge allowed attackers to mint rsETH tokens that were not actually backed by real ETH. These unbacked tokens were then deposited as collateral on Aave’s V3 markets, and the attackers borrowed large amounts of other digital assets (eg, ETH) against that collateral. As result Aave found itself sitting on a pool of “bad debt” worth hundreds of millions of dollars.
The market reaction was swift. Aave’s reported total value locked (TVL) dropped by around six billion dollars (!) almost overnight, even though the direct exploit only hit the rsETH‑related plumbing external to Aave itself. Depositors on Aave saw sky‑high utilization rates in the affected pools, which briefly made it difficult or impossible to redeem assets that normally appeared “safe.”
Since then, the Aave community and a broader coalition of DeFi protocols, including Compound and a group called “DeFi United,” have published a restoration plan aimed at recovering the missing backing ETH and ensuring that users who lost access to their funds will, in principle, get them back in full. That plan relies on commitments of real ETH from other actors in the ecosystem to plug the rsETH‑backing gap, rather than diluting unrelated depositors. If it works, the Kelp episode will become a case study in how DeFi can absorb a cross‑chain shock without fully socialising losses among ordinary users. [cryptobriefing](https://cryptobriefing.com/aave-tvl-plummets-6b-after-kelp-dao-hack-exploits-layerzero-bridge-flaw/)
#Lawyers, Of Course
This salutary tale of DeFi protocols and pools is interesting in itself, but there is in many ways an even more interesting parallel story developing in the courtroom. LayerZero, the cross‑chain messaging infrastructure used by Kelp DAO. has explicitly attributed the exploit to the Lazarus Group, via its “TraderTraitor” unit, a perspective shared with other industry voices. Crypto‑loss tracking firms and DeFi‑focused publications list the Kelp exploit among a cluster of recent hacks that they describe as linked to Lazarus, noting that the same group is also tied to the Drift Protocol exploit around the same period. Analysts point to familiar patterns—such as tunnel‑through‑bridges, laundering via known mixing services, and North‑Korean‑linked money‑laundering infrastructure—to reinforce the attribution.
Give that attribition, a number of aw firms have begun filing civil actions or motions aimed at recovering cryptocurrency allegedly stolen by North Korean‑linked hacking groups, most notably the Lazarus Group. In some cases, these suits are brought on behalf of victims of earlier historical incidents—such as the infamous 2000 “Reverend Kim” abduction case—leveraging old default judgments to argue that later‑stolen crypto ought to be treated as recoverable assets traceable to the same actors.
You can see their logic. Once legal authorities identify wallet addresses connected to North Korean cyber‑theft, those addresses become “known bad actors” on the chain. Lawyers can then seek court orders to seize assets that have flowed through those addresses or related counterparties, such as centralized exchanges or bridge operators accused of knowingly allowing the monetization of hacked funds. For example, a recent class‑action complaint alleged that a stablecoin issuer and its associated bridge were effectively letting North Korean‑linked attackers offload hundreds of millions of dollars in stolen crypto without properly freezing or blocking suspicious flows.
These cases sit at the intersection of sanctions law, antiterrorism statutes, and traditional property‑law concepts. And that is where the legal distinction between *fungible money* and *non‑fungible property* begins to matter.
##Fungible money versus non‑fungible property
Which brings us on to one of my favourite topics, fungibility. In common‑law systems, the treatment of assets depends crucially on whether they are fungible or non‑fungible. Fungible assets (eg, Dollars) are defined by kind and quantity rather than by individual identity. One dollar is interchangeable with another of the same denomination, whereas non‑fungible assets—such as a specific painting, a particular piece of land, or a unique NFT—are individualized and identifiable by their specific characteristics.
This distinction affects how the law thinks about tracing. If stolen cash is put into a bank account and then used to buy different assets, the claimant may have a claim against the new assets, but the analysis is highly fact‑specific and limited by the fungible nature of the money. In contrast, a non‑fungible asset can be more straightforwardly identified and returned because it is unique; you can point to a specific car or a specific artwork and get it returned to the rightful owner.
For non-fungible things, then, the law (in England) recognises two different kinds of property. The first is “things in possession”, which means broadly speaking assets such as a bicycle or a gold bar. The second is “things in action”, which means property that can only be claimed or enforced through legal action or proceedings, such as debt or shares in a company.
The Law Society proposed to add a third category to allow for what they call idiosyncratic objects. I might be tempted to label this category “things in wallets” but they have chosen the more generic “data objects” rather than, for example, “tokens”.
##What this means for crypto‑recovery lawsuits
The distinction between fungible and non-fungible matters in the crypto world. People might think about a bitcoin being interchangeable with another bitcoin, but the fact is that each bitcoin has a unique history. So while bitcoins might in certain circumstances display characteristics that enable them to function as a money substitute, they are not money.
In the Lazarus cases, lawyers are trying to apply this distinction. While one unit of ETH on a given chain is economically interchangeable with another, the fact that it is on‑chain data also makes it possible to trace particular coins. Every transaction is recorded; every wallet can be inspected; every hop from one address to another is visible. Hence the lawyers argue that the ability to identify “tainted” coins that passed through provably bad‑actor wallets and therefore should be treated as recoverable property rather than mere mixed‑with‑the‑world‑at‑large money. In practice, that means seeking court orders to freeze or claw back assets that can be shown to have flowed from sanctioned addresses or counterparties that allegedly helped the hackers convert those assets into usable value.
Absent a new body of law around data objects, if US courts accept that reasoning, they would be effectively treating cryptocurrency less like cash and more like monogrammed bars of gold whose serial numbers were recorded at every handover. That would make it easier to recover stolen funds far down the chain, even if they have changed hands several times which, as I have long argued, mitigates against using (for example) Bitcoin in commercial transactions that demand final settlment.
On the other hand, if US courts treat crypto strictly as ordinary fungible money, the tracing bar is much higher. Once the stolen coins are mixed into a larger pool—say, deposited on an exchange or swapped into other tokens—the claimant may have little more than an claim against the thief or a secondary actor, rather than a proprietary claim against the cryptocurrency.
**Josie Comment Here?***
#Signals and Noise
The Kelp‑Aave episode and the North Korean‑linked lawsuits together send a powerful signal to the wider industry. Technically, the Kelp hack is a reminder that DeFi protocols are only as safe as their weakest external dependencies—bridges, tokens, and oracles—and that even “non‑hacked” platforms can be exposed to massive indirect risk. Legally, the recovery‑focused lawsuits show that regulators and plaintiffs are no longer content just to track the math on the blockchain; they want to remobilize long‑standing common‑law distinctions about money and property to turn identifiable, traceable crypto flows into recoverable assets.
For lawyers, exchanges and the DeFi protocols themselves, the takeaway is that the line between technical custody and legal ownership is tightening with implications for the burgeoning stablecoin economy. I am very curious about the distinction between fungible central bank liabilities and non-fungible stablecoins backed by those liabilities. In a world where every coin can be traced, the traditional safety of fungibility may begin to erode—especially for assets that are openly linked to sanctioned actors. And that, in turn, is reshaping how the industry thinks about compliance, freezing mechanisms, and the very nature of digital property.
***Charles Kerrigan Here?***
Snippings 