When a Smart Contract Goes Wrong: Disputes, Jurisdiction and Remedies

Why disputes still happen

A smart contract is code deployed on a blockchain that automatically carries out agreed steps when certain conditions are met, such as releasing payment when a delivery is confirmed. The appeal is that execution does not depend on a person choosing to act. But automation does not remove the possibility of things going wrong. Code can contain bugs. The data feeding it, often called an oracle, can be wrong or manipulated. One party can argue the code did not reflect what was actually agreed. None of that disappears just because performance happened automatically, and England and Wales courts have shown they are willing to treat disputes arising from smart contracts in essentially the same way as disputes arising from any other contract.

The code is not automatically the whole agreement

A common misconception is that if the code executed, the matter is settled. In practice, courts look at the underlying agreement between the parties, which may exist in natural language alongside the code, or may need to be inferred from conduct, messages and the circumstances of the deal. Where the code and the apparent commercial intention diverge, for example because of a coding error that pays out the wrong amount, a court can still find that the parties’ true agreement differed from what the code actually did, and can order correction or restitution accordingly. The Law Commission’s work on digital assets and smart legal contracts, carried out at the request of the UK government, confirmed that existing contract law principles are generally flexible enough to accommodate this without new legislation, though it recommended some clarifying steps for the industry.

Jurisdiction and which court hears it

Blockchains are decentralised and often pseudonymous, which raises a practical question before any legal argument starts: which country’s courts even have authority to hear the dispute, and against whom. English courts have developed tools to deal with this, including granting orders against persons unknown where a wrongdoer cannot be identified by name, and permitting service of legal documents by unconventional means, including directly onto a blockchain address in some reported cases. Jurisdiction clauses written into the surrounding agreement, where one exists, usually settle the question in advance. Where there is no such clause and the parties are in different countries, the court will apply ordinary private international law rules to decide whether England and Wales is the appropriate forum.

Freezing and recovering crypto-assets

One area where the courts have had to adapt is interim relief. If a smart contract has misfired, for example sending funds to the wrong wallet or to a counterparty accused of fraud, a claimant often needs to act fast to stop the assets moving again before the substantive case is heard. English courts have granted freezing injunctions over crypto-assets and made orders requiring exchanges to disclose account information, treating certain crypto-assets as property capable of being protected in this way. This followed recognition, including through law reform work, that crypto-assets can fall within a third category of property distinct from the traditional categories of tangible things and rights that must be pursued through legal action, which matters for how they can be frozen, traced and recovered.

Oracles, bugs and who bears the loss

Many smart contracts rely on an oracle, a service that feeds real-world information, such as a price or a delivery confirmation, into the blockchain so the code knows when to act. If the oracle is wrong, manipulated or simply fails, the smart contract will faithfully execute on bad information. Liability in that situation depends on the contractual arrangements around the smart contract, not the code itself. Questions that typically arise include whether the oracle provider gave any warranty about accuracy, whether the parties allocated this risk in their agreement, and whether the error amounts to a mistake that can unwind the transaction. A poorly drafted smart contract that never addressed oracle failure, or counterparty default, or dispute resolution at all, leaves these questions to be answered under general contract and equity principles, which can produce outcomes neither party expected.

Remedies are still conventional

Even where a smart contract has executed exactly as coded, the available remedies if that outcome turns out to be wrong are the familiar ones: damages, rescission, rectification of the agreement, or restitution of a mistaken payment. Courts cannot reach into a blockchain and reverse a transaction directly in the way a bank might reverse a card payment, so remedies are usually aimed at the people involved rather than the code itself, for example ordering someone to transfer assets back or pay compensation. This is one reason lawyers advising on smart contract projects spend as much time on the surrounding legal documentation, dispute resolution clauses and governing law as on the code, since the code alone cannot provide a mechanism for resolving a disagreement about what it was supposed to do.

What this means in practice

For businesses and lawyers working with smart contracts, the practical lesson is that automation changes how an agreement is performed, not whether ordinary legal protections and risks still apply. Anyone entering into a significant smart contract arrangement should keep clear natural-language documentation of what was agreed, address known risks such as oracle failure and coding errors expressly, and specify governing law and jurisdiction, since none of that is supplied automatically by the code itself. Readers dealing with a live dispute involving crypto-assets or smart contracts should seek specialist legal advice given how fast the case law and regulatory guidance in this area continues to develop.