Key takeaways
- Evercrest Technologies, the company behind KelpDAO, has filed suit in British Columbia against LayerZero Labs, its Canadian affiliate and CEO Bryan Pellegrino over April's $292 million rsETH exploit, alleging negligent misrepresentation, negligence and defamation.
- The claim states Kelp users have withdrawn more than $650 million since the attack; reporting puts total outflows from LayerZero at nearly $15 billion. Pellegrino has publicly rejected the characterisation.
- The underlying question is unresolved across the whole industry: when a cross-chain message is accepted and value moves wrongly, liability sits somewhere between the messaging layer, the application that trusted it, and nobody at all.
- For anyone using cross-chain products, the practical takeaway is about exposure rather than blame: a position held through a bridge carries the security assumptions of every layer beneath it, not just the one you signed with.
Evercrest Technologies, the company behind KelpDAO, has filed a claim in British Columbia against LayerZero Labs, its Canadian affiliate and chief executive Bryan Pellegrino, over the roughly $292 million rsETH exploit in April. The claim alleges negligent misrepresentation, negligence and defamation, and seeks aggravated and punitive damages. Pellegrino has publicly rejected how the events are being characterised.
Most coverage of this will be about the money, the personalities, and who said what afterwards. The more useful thing in it is a question the industry has been avoiding for about five years, and this case forces into a courtroom.
What a cross-chain messaging layer actually does
Blockchains cannot read each other. A contract on Ethereum has no way to know, on its own, what happened on another chain. There is no shared clock and no shared state.
So a messaging layer sits in between. Something observes chain A, asserts to chain B that an event occurred, and chain B's contract acts on that assertion. Different designs make that assertion in different ways — a set of signers, a light client, a proof, a committee — but the shape is always the same. A claim is made, and a contract chooses to believe it.
That is the whole design space, and it is where essentially every large cross-chain loss in this industry has happened. Not in the cryptography, and rarely in the token contracts. In the moment where one system decides that a statement about another system is true.
Why liability here is genuinely unsettled
Consider the layers involved when value moves wrongly across chains.
There is the messaging protocol, which delivered a message. There is the application, which configured that protocol, chose its security settings and decided what to do on receipt. There is the token or vault contract that executed the resulting instruction. And there is the user, who interacted with a front end and in most cases understood none of the above.
Each layer has a defensible argument. The messaging layer can say it delivered exactly what it was asked to deliver, under settings the application selected. The application can say it relied on documented guarantees and marketing that described the layer as secure. The user can reasonably say they were given no meaningful way to evaluate either.
This is why the negligent misrepresentation allegation is the interesting part of the filing rather than the negligence one. It is not primarily a claim about code. It is a claim about what was said about the code — how the security model was described to the teams building on it.
That is a question courts are actually equipped to answer, and one the industry has never had answered.
The $15 billion detail
Reporting indicates nearly $15 billion has moved off the protocol, and the claim states Kelp users have withdrawn more than $650 million since the attack.
Outflow numbers in this industry need care. Capital moving off a protocol is not the same as capital being lost, and much of it reflects positions being unwound or relocated rather than value destroyed. Large notional figures also flatter themselves, because the same underlying assets can be counted at several layers of a stack.
What the number does indicate reliably is a change in willingness to hold exposure through that path. That is a real signal, and it is the one thing an outflow figure measures honestly.
What this means if you use cross-chain products
Not advice about any position. A structural point that applies regardless of what you hold.
When you hold a wrapped, restaked or bridged asset, you are not holding the underlying asset. You are holding a claim whose validity depends on every layer beneath it continuing to work correctly — the vault, the token contract, the messaging layer, and the configuration choices somebody made in that messaging layer on your behalf.
You signed a transaction with one application. Your risk is the union of all of them.
That is not an argument against using such products, and this kind of infrastructure exists because moving value between chains has genuine utility. It is an argument for knowing how many layers deep a position actually sits, because that number is almost always higher than the interface suggests.
Three questions worth being able to answer about anything cross-chain you hold: what asserts that the other chain's event happened, who can change that assertion mechanism, and what happens to your claim if the layer beneath it stops functioning.
Why a lawsuit, and why Canada
Worth noting because it reflects a broader shift.
For most of this industry's history, a protocol loss was resolved — or not resolved — socially. A post-mortem, a negotiation, sometimes a partial reimbursement, occasionally nothing. Litigation was rare, partly because it was unclear who to sue and in what jurisdiction.
That is changing as protocols acquire identifiable corporate entities in identifiable countries. Once there is a company with an address, ordinary commercial law becomes available, and the filing here names a Canadian affiliate and an individual executive alongside the main entity.
The consequence is not a moral one. It is that documentation matters considerably more than it used to. Descriptions of security guarantees written for a marketing page can end up being read as representations relied upon by a counterparty, and that is a different standard than the one most protocol documentation was written to meet.
The market
Bitcoin is around $83,900, effectively flat on the day and up roughly 3% over the week. Ether is near $2,683. Solana is around $120, having held above that level for the first time since January, with roughly 8% added over the week. XRP is the weakest of the majors on the day, off about 3.5%, though still up over the week.
Separately, Kalshi has lost its appeal, setting up a potential Supreme Court case — a thread worth following for anyone interested in how prediction markets get classified, which is a question with direct implications for onchain equivalents.
What to take from this
A very large exploit has become a commercial lawsuit, and the allegation that matters most is about what was said rather than what was coded.
The question underneath it — who is responsible when one chain wrongly believes something about another — has no settled answer anywhere in this industry. Until it does, the only reliable protection is knowing how many trust assumptions sit under a position, because no court outcome will retroactively reduce that number.
Frequently asked questions
What exactly is LayerZero being sued over?
Evercrest Technologies, the company behind KelpDAO, has filed a claim in British Columbia against LayerZero Labs, its Canadian affiliate and chief executive Bryan Pellegrino relating to April's roughly $292 million rsETH exploit. The allegations are negligent misrepresentation, negligence and defamation, and the claim seeks aggravated and punitive damages. Pellegrino has publicly rejected the characterisation of events. Nothing has been determined by a court.
Why is 'negligent misrepresentation' the notable allegation?
Because it is a claim about what was said rather than what was built. Negligence claims about smart contract behaviour require a court to evaluate code and industry standards that barely exist. Misrepresentation asks a more familiar question: was the security model described in a way that another business reasonably relied upon. Courts answer that kind of question routinely, which is why this case could produce a usable precedent where purely technical claims often do not.
Does nearly $15 billion leaving mean $15 billion was lost?
No. Outflows measure capital being moved or unwound, not destroyed. The reported loss in the incident is around $292 million. Large notional figures in cross-chain infrastructure also tend to overstate distinct value, because the same underlying assets can be counted at multiple layers of a stack. What an outflow number does show honestly is reduced willingness to hold exposure through that route.
How do I tell how risky a cross-chain position is?
Three questions. First, what mechanism asserts that an event on the other chain actually happened — a signer set, a committee, a proof, a light client. Second, who has the power to change that mechanism or its settings, and how quickly. Third, what happens to your claim if the layer beneath it stops functioning. If those cannot be answered from public documentation, that is itself the answer about how much is being taken on trust.
Is cross-chain infrastructure inherently unsafe?
It carries a category of risk that single-chain activity does not, because it requires one system to accept a claim about another, and that acceptance step is where most large cross-chain losses have occurred. That is a structural property of the problem rather than a flaw in any one design. It does not make such infrastructure unusable — moving value between chains has real utility — but it does mean the risk is additive across every layer involved, not limited to the application you interacted with.
Not financial advice. Crypto assets are volatile and unregulated in many jurisdictions. In India, gains are taxed at 30% with 1% TDS on transfers. Do your own research and never invest money you cannot afford to lose.
Editorial note: Crypto Shakti uses an AI-assisted research and drafting workflow. Every article is grounded in the linked primary sources and live market data captured at publication time.