Chris Wray started the session by giving an overview of distributed ledger technology, blockchain and smart contracts.
He ran through issues with running smart contract code in one of its main use cases, payment systems, including:
- Relying on real-world data (or ‘trusted indices’) without a trusted provider of that data;
- Transactions that are essentially public; the use of blockchain as an automated timestamp / notary application could lead to unexpected commercial exposure through the metadata;
- The lack of an inherent solution to the problem of the ‘real-world’ identity of parties – a problem from an AML-perspective;
- The extreme volatility of various cryptocurrencies with respect to real-world assets such as USD;
- The irrevocability of transactions. For example, in the case of an accidental transfer of funds;
- Private-key risk. Either a trusted intermediary holds the key for you or the individual does. Both scenarios have their own security risks.
There was discussion over how advanced escrow conditions can ever be under a smart contract and the more fundamental question, especially with regard to financial instruments, of whether smart contracts are necessary at all.
David Mcllwaine then discussed the topic of smart contract disputes. If a remedy has not been anticipated within a smart contract, the answer can be determined by a form of dispute resolution which can either be:
- “off-chain”, in which arbitration is the preferred method, not litigation; or
- “on-chain”. Examples such as Kleros, Juris, Codelegit, and Confideal were given in which the resolution and the enforcement of the decision happen on the blockchain.
Potential problems with smart contracts were then examined, including potential coding errors, hacking and cybersecurity. Specific legal concerns include contract validity under different jurisdictions and procedural issues including the availability of evidence, of assets and enforceability.
Juris was then looked at as an example of an “on-chain” arbitration protocol. The parties adopt the Juris code in their smart contract. In the case of a dispute, the system freezes the contract and moves funds to a safe wallet. The system details the complaint and if the parties have not settled the dispute it offers a three-stage resolution from a standard mediation up to a binding arbitration award.
Sarah Compani then described the dispute resolution protocol, Kleros, an Ethereum-based project utilising game theory.
Jurors are selected by holding tokens and make decisions in favour of one party. The minority jurors forfeit their tokens to the majority jurors, thereby paying for the process.
However, it was noted that laws of various jurisdictions and the concept of a ‘fair trial’ are regularly implied in decisions made by jurors, even when they are not specifically included within the smart contract. Disproportionate levels of influence can also be given to various jurors based on their apparent authority.
Challenges for regulators were then discussed, including the issue of how to ascertain the correct jurisdiction for computer code where parties are international and smart contracts are not tied to a single legal system.
GDPR also raises problems as smart contract information should not be publicly available and cannot be retained indefinitely.
Other challenges include:
- The identity of the parties, which can be anything from a person, an oracle or even a bot;
- Proof that a contract actually existed in the first place;
- Proof that the parties intended to be legally bound by the smart contract;
- Enforceable decisions under the Kleros protocol may not be valid at law;
- The potential need for additional, off-chain formal contracts (e.g., handwritten acknowledgements).
The audience then discussed the issue of potential judicial interference if the parties are not satisfied with the on-chain decision.
Even in a conventional court, there would then be questions over:
- The remedies available, which are severely limited (e.g., no injunctions);
- The potential difficulty of providing restitution for something that should be unique (for example, a cryptocurrency) and whose value fluctuates hugely over time;
- Dealing with third parties who don’t recognise its authority.
David Quest QC then discussed the challenges in defining the legal status of cryptoassets under English law, which is to be looked at shortly as part of the UK Jurisdiction Taskforce’s Legal Statement, of which he is an author.
Amongst the legal challenges being wrestled with is whether cryptocurrency is a new class of asset or isn’t an asset at all. If Bitcoin can’t be classed as one type of property (‘chose in possession’), then it either has to be classed as the other (‘chose in action’) or it can’t be said to be property. There is also the conceptual question of whether cryptoassets behave in the same way as traditional legal property.
David looked at several hypothetical scenarios to illustrate the ambiguity surrounding the legal status of cryptoassets under English law.
Among the questions arising from these scenarios were:
- Who is the actual owner a cryptocurrency? What happens when a private key is given to someone else? Would that be a transfer of ownership or is it simply split? The private key creates an ability to control the asset, or the potential ability to sign a transfer. But whether or not that is subsequently accepted on the blockchain is a completely different question.
- Can a trustee in bankruptcy take control of cryptoassets as part of the insolvency estate? Most of the audience agreed that there was no real reason to treat cryptoassets differently from any other sellable asset in this context.
- X uses a private key to transfer the same cryptoasset to A and B. One is accepted in the distributed ledger, the other not. What rights do A and B both have in relation to the cryptoasset? It seems natural that the order of priority should follow the record of the blockchain. General consensus of parties using the blockchain is that the first party onto the chain should be the correct one (the chain being the ultimate arbiter of truth).
- However, the law isn’t bound by the consensus of the blockchain. Traditionally, the earlier assignment/transfer is the authentic one (subject to various questions about perfection of title), and after such a point the seller X no longer has title over the property in order to give it away.
- There is merely an ‘agreement to transfer/assign’ the private key, but it is only the consensus that affects the transfer at the stage that it is placed onto the blockchain. The consensus transfers the asset. But the aggrieved party will still have a contract claim against X for failing to actually transfer the asset.
Matthew Lavy, another author of the forthcoming Legal Statement, then looked at the challenges of dealing with smart contracts under English law.
Matthew gave the example of the 2016 DAO, a venture fund for smart contracts. Because of an ‘erroneous’ mechanism, 60m USD was hacked. Was this a bug or a feature of the program? Can the principle of rectification (a creature of equity) be relied upon to legally change the actual software code (in this case, the contract)?
He then looked at several hypothetical scenarios to illustrate different challenges:
- What if multiple smart contracts are individually negotiated, coded and verified but together create an unintended consequence
- What if there is a bug on the development (execution) platform, not in the smart contract itself, and the platform is open source? There would be no other third-party cause of action and even though the contract terms weren’t executed properly, the original contractual obligation would still remain.
- The identities of parties to a contract affect the nature of the contract itself, either through interpretation or through the inclusion of various (otherwise absent) implied terms (e.g., the Consumer Regs implied terms in any English consumer contract).
- If smart contract usage increases and there is a shift away from exclusively sophisticated users (and subject to the jurisdiction question), how can statutory requirements such as these ever be satisfied without the use of party identifiers?
- And without verifiable identifiers, is it really conceivable that the nature of a contract can only be defined at the dispute resolution stage where an identity is revealed out of necessity?
Allan Murray is a Senior Associate at iLaw, London
