Arbitrating disputes arising out of smart contracts
What are smart contracts?
One of the most exciting aspects of distributed ledger technology (DLT)[1] is the development of so-called ‘smart contracts’. The term ‘smart contract’ encompasses several different concepts and it is therefore difficult to formulate a single, precise definition of what is meant by the term. Nevertheless, the essence of a smart contract is that it is self-executing. In other words, a smart contract will “automatically execute certain tasks once pre-defined conditions are met, and the execution of which cannot be halted unless this possibility is contemplated by the relevant parties and embedded within the relevant computer code”[2].
An analogy that is commonly drawn is that of a vending machine. The consumer selects the drink that he or she wishes to purchase by pressing a button on the outside of the machine. As soon as money is inserted into the machine, an irreversible transaction is initiated, whereby the consumer will automatically receive the drink selected and the money cannot be recovered. There are many contracts that operate according to a similar logic. One example that has been given is that of a flood insurance policy, linked to a feed of precipitation data, whereby once the data feed shows that precipitation has reached a certain level, the policy pays out[3]. Other examples might include contracts involving royalty payments, escrow agreements and mortgage contracts.
What are the advantages of smart contracts?
The primary advantage of smart contracts lies in the automation of execution and the increases in efficiency / reduction in costs that accompany such automation. If a smart contract runs correctly, there should be no need to involve the Courts or an arbitral tribunal in order to seek enforcement of the contractual terms because the terms are automatically enforced.
Automated contracts are not new. However, before the proliferation of DLT, both parties in practice needed to program their own computers, run separate instances of the program on their systems and code their own versions of the relevant provisions. This would be particularly necessary given that “one party might not be willing to rely on the other party’s code, and so each would code their own version of the relevant provisions”[4]. This not only involves a risk that two implementations of the code do not match, but may also allow for the possibility that one party could tamper with its version of the code after it had been switched on. In the absence of DLT, the other alternative for the parties would be to involve a third party to run the code, which would require both parties to trust that third party and would involve an extra layer of administrative burden and costs.
DLT allows the smart contract code to be embedded in the distributed ledger, such that one single version of the code exists that binds the parties and does not require the involvement of a trusted third party. Moreover, once the code has been switched on, “the parties can take comfort from the fact that…neither party can tamper with [it]”[5].
Are Smart contracts actually contracts?
There are those who argue that ‘smart contracts’ are not contracts at all in any real sense, but are simply computer programmes which can be used to guarantee performance[6]. There is merit in this view, for no legal relationship exists in a vacuum. A smart contract will only be considered as a contract by a national court or arbitral tribunal if it possesses all the constituent elements of a contract under the relevant national law. In England and Wales, the Law Society’s ‘LawTech Delivery Panel’ is to address whether or not English law considers that a smart contract is capable of giving rise to legally binding obligations in its Legal Statement to be released on 18 November 2019. The consultation paper provides that “this is likely to involve the application of established tests under English law for determining whether a binding legal contract arises. The concepts of offer, acceptance and consideration are likely to be relevant in this context”[7].
The answer may, in fact, depend on the smart contract implementation in question. Whilst some smart contracts are written solely in computer code, others are written in computer code alongside a parallel natural language contract. It would be difficult to argue that a binding legal contract does not arise in circumstances where the computer code is supported by a natural language version. As one commentator has written, “at present, the only way we see to be sure that the output or performance of smart contract code has the force of legal obligation is to confer that legal binding effect by means of a more traditional legal contract that serves as a legal “wrapper” for the self-executing code”[8].
Disputes involving smart contracts
One of the misconceptions around smart contracts is that, because they self-enforce, there is no need to consider dispute resolution mechanisms because, quite simply, disputes will not arise. As the Chancellor of the High Court of England and Wales has said, “many of the coders, who are developing the algorithms for smart contracts, tend to believe, everywhere – not just in the UK – that no legal basis is necessary because the answer to every dispute is built into the code”[9]. As a dispute resolution lawyer, it is not difficult to identify possible scenarios in which disputes may arise, including but not limited to:
i. One party may allege that the other party made a fraudulent and/or negligent misrepresentation as to the effect of the smart contract;
ii. The code may not accurately reflect what the parties understood to be their agreement (e.g. common mistake of law or fact);
iii. The smart contract could be void for illegality;
iv. One party may not have legal capacity to enter into the smart contract (e.g. by being under the legal age);
v. The smart contract may contain coding errors meaning that it is not performed and/or is incorrectly performed.
In these cases, parties may well need to resort to a dispute resolution mechanism to remedy the position.
Arbitrating smart contract disputes
Arbitration has certain features that naturally lend it to resolving disputes arising out of smart contracts.
In the first place, smart contract disputes will often involve witness and expert evidence containing information that the parties wish to keep confidential relating to new technology and/or software. In many jurisdictions, including England and Wales, there is an implied duty of confidentiality on the parties, which can often mean that documents disclosed or used in the arbitration cannot be disclosed or used for any other purpose. This is likely to be an advantage to those parties who do not want their proprietary information to become public.
Secondly, as smart contracts using DLT operate via a distributed network of computers, it will often be difficult for a national court to make a determination as to whether or not it has jurisdiction to decide a dispute and/or what the governing law of the smart contract is (for example, by reference to common metrics such as place of performance). An agreement to arbitrate could eliminate this uncertainty and the risk of satellite disputes. An alternative approach may be to consider that, since smart contracts are inherently ‘delocalised’ in that they operate via a distributed network of computers, a ‘delocalised’ approach to smart contract arbitration is preferable, such that smart contract arbitrations should not be tied to any one national legal system.
Finally, depending on whether or not the parties have selected ‘on-chain’ or ‘off-chain’ arbitration, it may be possible for the award to benefit from an automated enforcement procedure (see below).
On-chain v off-chain arbitration
Broadly speaking, the possible approaches to arbitrating smart contract disputes fall into two camps, namely ‘on-chain’ and ‘off-chain’ arbitration.
On-chain
‘On-chain’ arbitration essentially involves the use of technological solutions by which the equivalent of a traditional arbitral award is automatically enforced by the smart contract. By way of example, “this may be done…by the parties making available to the smart contract certain assets (e.g. cryptocurrencies) which, upon occurrence of a defined condition (issuance of an “award”), are transferred from one party to the other”[10].
It is conceivable (although no such solution has yet been effectively implemented) that ‘on-chain’ arbitration could occur in an entirely robotised manner, without requiring the involvement of any action taken either by a third party or by the parties themselves. For example, it has been suggested that an “algorithm could resolve the dispute based on the analysis of similar transactions and disputes”[11]. The reality, at least for the foreseeable future, is that there is not yet the technology to provide for an entirely automated system of arbitration, in which both the legal and technology communities can have confidence.
Perhaps a more realistic ambition in the short term is for a partially automated system. In this instance, the smart contract could again provide for the automatic transfer of assets to the other party upon the issuance of an award, but the award is not rendered by way of robotised algorithm, but rather by human arbitrators, following a traditional arbitral process involving the presentation of evidence and submissions, whose award could be fed back onto the blockchain and automatically executed. A number of ventures have emerged in recent years along similar lines, but are yet to garner widespread adoption[12]. At present, these solutions may find use in relation to simple contractual disputes, where the scope and nature of a possible arbitral award can be easily foreseen at the time that the smart contract is formulated. However, there is still some way to go before these solutions are likely to be implemented in more complicated contractual scenarios.
Off-chain
Off-chain arbitration is more akin to the ordinary arbitral process in that it lacks the automatic enforcement of the award. This leads to further complications because Article II of the New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards 1958 requires an agreement to arbitrate to be “in writing”. It is uncertain whether an award issued pursuant to an arbitration agreement contained in the code of a smart contract would be capable of being enforced. This is perhaps another reason why parties should give serious consideration to whether or not to enter into a parallel natural language contract.
Moreover, given that smart contracts, once the relevant conditions are met, set in motion an irreversible transaction, arbitral tribunals may need to find creative remedies in their awards that are capable of resolving disputes fairly. It may be that such remedies would be facilitated by specialist smart contract arbitrators and a set of arbitral rules designed specifically to deal with smart contract disputes. Interestingly, adjudication procedures with specific rules designed to deal with technology disputes have been developed in recent months[13].
Conclusion
It is plausible that, as blockchain technology continues to develop, smart contracts will become more and more widespread. It is vital that legal systems adapt in order to provide a proper legal foundation in which they can properly operate. This includes modern systems of dispute resolution and, given its obvious advantages, arbitration. The extent to which arbitration can become automated is, at present, unclear and may involve some trial and error over the coming years. What is clear, however, is that those involved in the technology sector and many of its investors, see the potential that automation based on DLT has for the international business community. The arbitration community should be wary not to lag behind.
Robert Coffey,
Managing Partner
Peter Stewart,
Associate (Employed Barrister)
Cooke, Young and Keidan, London
[1] A distributed ledger is a database of replicated, shared, and synchronized digital data that is geographically spread across multiple sites in a network. Source: https://medium.com/blockchain-review/whats-the-difference-between-blockchain-distributed-ledger-technology-19407f2c2216.
[2] Public Consultation – the status of cryptoassets, distributed ledger technology and smart contracts under English private law, UK Jurisdiction Taskforce of the LawTech Delivery Panel, p. 30 (May 2019).
[3] https://www.nortonrosefulbright.com/en/knowledge/publications/ea958758/arbitrating-smart-contract-disputes.
[4] https://www.isda.org/a/6EKDE/smart-contracts-and-distributed-ledger-a-legal-perspective.pdf p. 9.
[5] Ibid.
[6] Cheng lim, T.J. Saw and Calum Sargeant, ‘Smart Contracts: bridging the Gap Between Expectation and Reality’
[7] Public Consultation – the status of cryptoassets, distributed ledger technology and smart contracts under English private law, UK Jurisdiction Taskforce of the LawTech Delivery Panel, p. 31 (May 2019).
[8] https://talkingtech.cliffordchance.com/en/emerging-technologies/smart-contracts/are-smart-contracts-contracts.html.
[9] https://www.judiciary.uk/wp-content/uploads/2019/05/CHC-speech-LawTech-Smart-Contracts-and-Artificial-Intelligence.pdf p. 2.
[10] https://newtech.law/en/on-chain-and-off-chain-arbitration-using-smart-contracts-to-amicably-resolve-disputes/.
[11] https://www.hoganlovells.com/~/media/hogan-lovells/pdf/2018/2018_12_13_icc_robots_arbitrator.pdf p. 29.
[12] See for example Kleros https://kleros.io/whitepaper.pdf and Mattereum https://mattereum.com/services/.
[13] https://www.scl.org/files/download/1836-SCL%20Adjudication%20Scheme%20Rules%20Version%201.pdf.