Observations on the UK Law Commission’s Call for Evidence on Smart Contracts
In November 2019, the UK Jurisdiction Taskforce (“UKJT”), a subsidiary of the UK’s LawTech Delivery Panel, published a Legal Statement on the Status of Cryptoassets and Smart Contracts (the “Legal Statement”). In our previous article entitled “UK Jurisdiction Taskforce Publishes Legal Statement on Status of Cryptoassets and Smart Contracts—Observations from Ireland”, we discussed the Legal Statement with reference both to its statements and its potential implications for those in Ireland with an interest in its subject matter.  Following the publication of the Legal Statement, the UK Government asked the Law Commission of England and Wales (the “Law Commission”) to undertake a scoping study into the law on smart contracts. On 17 December 2020, the Law Commission published a call for evidence on smart contracts as a first step in this scoping study (the “Call for Evidence”). The Call for Evidence is currently seeking views about, and evidence of, the ways in which smart contracts are used and the extent to which existing English law can accommodate smart contracts.
In this article, we discuss the Call for Evidence with reference both to its statements and its potential implications for those in Ireland with an interest in its subject matter. To ensure that Ireland remains a competitive choice for legal services and dispute resolution post-Brexit, there is a compelling case for reviewing the current legal framework in Ireland to ensure that it facilitates the use of smart contracts, including matters of creation and enforcement.
Scope of Call for Evidence
The Call for Evidence covers the following topics:
- what is a smart contract?;
- formation of smart contracts;
- interpretation of smart contracts;
- remedies and smart contracts;
- consumers and smart contracts; and
What Is a Smart Contract?
A smart contract was first described in 1996 as a “computerised transaction protocol that executes the terms of a contract”. For the purposes of its Call for Evidence, the Law Commission defines a smart contract as “a legally binding contract in which some or all of the contractual obligations are recorded in or performed automatically by a computer program deployed on a distributed ledger.” Smart contracts can be used to record and perform the obligations of a legally binding contract and are known as smart legal contracts. A distinguishing feature of a smart contract is that some or all of the contractual obligations can be performed automatically. The world is already accustomed to automatic contracts, for example, automatic bank transfers and purchasing products when engaging in online shopping. However, as these transactions still require some human interaction (e.g. delivery of goods) or include third parties (e.g. a bank), they do not amount to “smart” contracts.
Types of Smart Contracts
The Call for Evidence identifies three types of smart contracts:
- a natural language contract where the contractual obligations are performed automatically by a piece of code. The code itself does not contain any of the contractual terms;
- a hybrid contract consisting of code and natural language; and
- a contract consisting solely of code. The Law Commission’s initial view is that this type of contract is likely to present the most challenges from a legal perspective, in terms of determining whether and when a smart contract is formed, and how that smart contract can, or should, be interpreted.
Distributed Ledger Technology (“DLT”)
DLT is being promoted as a way to deploy smart contracts and involves contractual obligations being expressed in computer code and performed by computers in a network. A DLT system is comprised of a digital database (ledger) which is shared among a network of computers (nodes). The most famous example of a DLT system is Bitcoin. The distinguishing feature of DLT compared to other shared databases is that the ledger is not maintained by a central administrator. Instead, the ledger is maintained collectively by the nodes and no individual node has the power to add data to the ledger. Nodes can propose new data entries, but entries will only be added to the ledger when all nodes reach a consensus that the entry should be recorded. This is known as the “consensus mechanism”. The mechanism is designed so that once data is added to the ledger, the possibility of amending such data lies somewhere between being extremely difficult and impossible to do, depending on the architecture of the particular DLT facilitating the smart contract. A smart contract is triggered by addressing a transaction to it, which is then executed automatically in a prescribed manner on every node in the network, according to the data included in the triggering transaction.
Formation of Smart Contracts
As noted by the UKJT in its Legal Statement, English law does not normally require contracts to be in any particular form. In our previous article, we confirmed that this is also true in Irish law. The general rule in Ireland is that a contract does not have to be in writing before it can be enforced. For example, in Pernod Richard & Comrie plc v FII (Fyffes) plc, an oral agreement for a multi-million-pound takeover was enforced. The decision was subsequently upheld on appeal by the Irish Supreme Court.
In its Call for Evidence, the Law Commission agrees with the Legal Statement and states that the requirements for formation of smart contracts are the same as normal contracts, namely that:
- agreement has, objectively, been reached between the parties as to terms that are sufficiently certain;
- the parties intended, objectively, that they would be legally bound by their agreement; and
- unless the contract is made by deed, each party to it must give something of benefit (consideration) because a gratuitous promise in return for nothing is not generally enforceable.
Offer and Acceptance
A legally binding contract must comprise an offer based on specified terms and an acceptance of those terms. There will be agreement if A offers terms to B, and B accepts those terms by words or conduct. Agreement is generally found in, or evidenced by, a written document bearing signatures of A and B but, as explained above, writing or signature is not a necessary precondition to the enforceability of a contract. Some negotiations between the parties (e.g. email, instructions to coders or oral conversations) concerning natural language will most likely precede smart contracts. This is similar to a traditional contract scenario where the words and conduct of the parties lead to the offer and acceptance of a contract.
Where there is no natural language communication between parties who have entered into a smart contract, offer and acceptance may be more difficult to determine. For example, A might deploy a piece of code on a distributed ledger and B might interact with that piece of code, causing the code to execute a transaction. Perhaps no natural language documents or communications were exchanged and the parties’ interactions may have consisted exclusively of transactions on a DLT mediated by the computer program. In this case, does the deployment of the code amount to an offer, or is this merely an invitation to treat?
In the English case of Thornton v Shoe Lane Parking Ltd, the defendant installed a machine in his car park that would automatically grant entry to the car park when a customer inserted money into the machine. It was held that the defendant, in holding out the machine as being ready to receive money, was making an offer to customers to use the car park in exchange for payment. The Call for Evidence argues that the same reasoning could apply to a smart contract, i.e. a person who deploys a computer program which will automatically transfer an asset upon receiving payment could amount to an offer and the payment of money could amount to acceptance.
In general, each party to a contract must give something of benefit (known as “consideration”). The “smart” nature of the contract, being the embedding of terms of the contract in a networked system that executes and enforces performance using various techniques such as the consensus mechanism discussed above, does not preclude A and B from giving each other something of benefit and therefore the issue of consideration is unlikely to present challenges in a smart contract context.
Intention to Create Legal Relations
Where a smart contract includes a natural language component, it is unlikely that there would be difficulty in proving that the parties intended to create legal relations. This is because, in the case of an express agreement made in a commercial context, an intention to create legal relations is presumed. However, difficulties could arise if the agreement is made as a result of an interaction on a distributed ledger where the agreement is inferred from the parties’ conduct. Would the presumption that the parties intended to create legal relations apply? If it is generally understood by the users of the DLT system that interactions on the ledger do not attract legally enforceable obligations, then this might be a factor weighing against finding an intention to create legal relations in this scenario.
The general rule is that contracts do not need to be in any particular form. However, some statutes require certain contracts to be made “in writing” and/or “signed”. Requirements that a contract be “in writing” or “evidenced in writing” are very rare in English law and indeed Irish law. However, they do exist. If the terms of a smart contract are set out in a natural language document, then the smart contract would likely satisfy an “in writing” requirement. However, where contractual terms are recorded in code, satisfaction of an “in writing” requirement may depend on whether or not the code is in a form that can be read by a natural person. Code is initially drafted in high-level programming language known as “source code”. Source code uses both words and symbols and can be read by coders. Source code is then compiled into machine-readable “object code” which is impossible to read, even by a coder. If the terms of a smart contract are contained in source code, it is arguable that those terms can be read by a natural person and the smart contract containing such source code would satisfy an “in writing” requirement. However, if the terms of the smart contract reside in object code, it may be difficult to argue that the contract is “in writing” because object code cannot be read by a natural person.
However, code is, no matter the difficulty or impossibility of being understood by natural persons, still “in writing”. Therefore, a question which may arise is whether or not a statutory “in writing” requirement is capable of being satisfied by code if such code is simply “in writing”, regardless of whether or not such writing is in fact understandable by a natural person. The Legal Statement noted that the mere fact that a smart contract is in electronic form does not mean that it cannot satisfy a statutory “in writing” requirement. The Legal Statement continued that the question may be whether there is something intrinsic to computer code, as opposed to human language, that should lead to a different conclusion. Time will tell.
What Does “In Writing” Mean?
According to Sch.1 to the Interpretation Act 1978 (the “1978 Act”) in England and Wales, “writing” is defined as including “typing, printing, lithography, photography and other modes of representing or reproducing words in a visible form, and expressions referring to writing are construed accordingly.” The UKJT’s view is that, to the extent the relevant code can be said to be representing or reproducing words and be made visible on a screen or printout, it is “likely to fulfil” a statutory “in writing” requirement. According to Pt 1 of the Schedule to the Interpretation Act 2005 (the “2005 Act”) in Ireland, “writing” is defined more broadly than under the 1978 Act and includes:
“printing, typewriting, lithography, photography, and other modes of representing or reproducing words in visible form and any information kept in a non-legible form, whether stored electronically or otherwise, which is capable by any means of being reproduced in a legible form”.
It is arguable that the words “and any information kept in a non-legible form, whether stored electronically or otherwise, which is capable by any means of being reproduced in a legible form” in Pt 1 of the Schedule to the 2005 Act are more capable of providing support for the proposition that code found within smart contracts can satisfy a statutory “in writing” requirement than the definition of “writing” in Sch.1 to the 1978 Act. This is because code is, to judges, lawyers and non-coders, in a non-legible form, but, critically, is capable of being reproduced in a legible form for comprehension by judges, lawyers and non-coders through the assistance of extrinsic evidence, expert evidence or exceptions to the parol evidence rule. It is arguable, therefore, that a statutory “in writing” requirement can be more easily satisfied by smart contracts composed partly or wholly of code under Irish law than under English law.
Can a Smart Contract Be Signed?
If the terms of a smart contract are recorded in a natural language document, the smart contract could be signed in the ordinary way. Where a smart contract consists solely of code, however, the parties may sign the contract electronically, for example, by using a digital signature to authenticate code deployed on a DLT system. In its Legal Statement, the UKJT stated that a statutory signature requirement is “highly likely” to be capable of being satisfied by using a private key, because an electronic signature which is intended to authenticate a document will generally satisfy a statutory signature requirement, and a digital signature produced using public-key cryptography is a particular type of electronic signature.
Irish law already accommodates electronic signing. Section 12(1) of the Electronic Commerce Act 2000 (the “2000 Act”) provides that if by law or otherwise a person or public body is required or permitted to give information in writing, then, subject to certain conditions in s.12(2), the person or public body “may give the information in electronic form, whether as an electronic communication or otherwise”. Section 13(1) of the 2000 Act provides that if by law or otherwise the signature of a person or public body is required or permitted, then, subject to certain conditions in s.13(2), “an electronic signature may be used”.
Interpretation of Smart Contracts
An important benefit of smart contracts over traditional contracts is the lack of textual ambiguity, which may reduce the need for lawyers to use anachronistic canons of construction and other textual interpretation techniques. When interpreting a contract, a judge asks himself or herself what the language would have meant to a reasonable person, equipped with all the background knowledge available to the parties at the time the contract was made. Where a smart contract is written in natural language, this will not cause any novel issues. However, where the terms of the contract are written in code, a judge will need to consider how to interpret a smart contract that he or she may be unable to read, and what the smart contract would have meant to a reasonable person, equipped with all the background knowledge available to the parties at the time the smart contract was made. One approach to interpreting a smart contract contained exclusively in code may be to ask what a computer would do upon receiving the coded instructions. In theory, this should produce a definite answer, but the retention of expert advisors and divergence of opinion on the make, model or operating system of this computer may lead to varied results.
Another option is to ask what a reasonable person with coding knowledge would do. An expert coder could assist the court by translating the code, similar to a language expert. However, it is unlikely that simply translating the code would be sufficient to assist a court. The coding expert would have to give his or her reasoned opinion as to what the code actually means or instructs a computer to do. Such an approach arguably shifts the role of interpretation from the judge to the coding expert. Furthermore, who, in law, is a reasonable person with coding knowledge? An expert coder is unlikely to be a reasonable person with coding knowledge.
In a hybrid contract containing natural language and code, there may be discrepancies between the natural language and the coded elements of the smart contract. Under traditional contract law, if there is a conflict between two terms in a contract, the court will look at the contract as a whole. Applying this interpretive approach in a smart contract context, a judge may have to look at the coded element and the natural language element of a smart contract. Since a judge, being a natural person, understands natural language, could there be a risk that the judge may be naturally inclined to prefer the term contained in natural language over the term contained in code? To avoid this scenario, the Law Commission in its Call for Evidence suggests that contracting parties could include an order of precedence in smart contracts where, for example, the interpretation of natural language will trump the interpretation of code in circumstances where there is a conflict between natural language and coded elements of a smart contract.
Even in situations where a smart contract is based entirely in code, it is likely there will have been some negotiations between the parties prior to entering into the smart contract. Ordinarily, pre-contractual information is inadmissible in law due to what is known as the parol evidence rule. However, the law allows exceptions. One exception allows the admissibility of pre-contractual information to resolve ambiguities in a contract or to assist in interpreting a term of the contract. This exception may allow admission of natural language information from pre-contractual negotiations between parties to a smart contract in order to assist a judge in interpreting the coded terms of a smart contract.
Remedies and Smart Contracts
Society for Computers and Law Adjudication Scheme
Contractual interpretation will likely be a key issue in any litigation or dispute resolution proceedings involving a smart contract. A court will likely require its own expert advisors, and parties may retain their own experts to interpret technical terms. Proceedings will likely be protracted, at least until a body of precedent is developed rendering only the most novel issues susceptible to contention. This will take time. One option available to parties contemplating entering into smart contracts may be to stipulate in such contracts that all disputes arising out of or in connection with the smart contract are to be adjudicated under the Society for Computers and Law’s Adjudication Scheme (“SCL Adjudication Scheme”). Due to the SCL Adjudication Scheme’s retention of a panel of expert lawyers and computer experts, its cap on hourly adjudicators’ fees, its guarantee of a decision within three calendar months and its preservation of the right to litigate or arbitrate a dispute within six calendar months of the fifth working day after a decision is delivered, the SCL Adjudication Scheme may soon become the optimal method for resolving smart contract disputes. For more information, please see our recent briefing on the SCL Adjudication Scheme (here).
Under contract law, a court can “rectify” a contract where it does not accurately reflect the parties’ agreement. In relation to smart contracts, if the parties engage a coder to translate certain elements of their agreement into code and this code does not accurately reflect what was agreed, rectification could be used to amend the contract.
A contract can be rendered “void” (meaning it has no effect from the beginning of the contract) if one or both of the parties laboured under a mistake when entering into the contract. For example, the parties may hold beliefs or assumptions about how the code will execute their agreement. If these beliefs or assumptions are mistaken, can a smart contract be void on grounds of mistake? Where both parties are mistaken about a matter relevant to the execution of the code (known as a “common mistake”), the common mistake will only render the contract void if it makes performance of the contract or the achievement of the purpose of the contract impossible. Where one party is mistaken about the execution of the code (known as a “unilateral mistake”), the unilateral mistake will only render the contract void if the mistake relates to a term of the contract and the other party was aware of the mistake at the time of contracting. A bar to the availability of this remedy is that it will be particularly difficult to prove a party’s knowledge of a mistake where a smart contract is entered into by computer programs on behalf of the parties.
A contract is “voidable” (meaning it is liable to be set aside from the beginning of the contract) if one party is induced to enter the contract by a misrepresentation made by the other party. Similar to normal contracts, the entering into of smart contracts will likely be preceded by a period of negotiation between the parties. Existing Irish and English law will be capable of determining whether a party to a smart contract made a misrepresentation, by their words or conduct, which induced the other party to enter the smart contract.
The Law Commission anticipates that restitutionary remedies will be particularly useful in the context of smart contracts. Some or all of the terms of a smart contract are performed automatically by code on a distributed ledger and there may be no mechanism for parties to stop the execution of the code upon discovery of a factor rendering the contract void or voidable. The code may continue to execute (and perhaps fully execute) the contract regardless of the fact that mistake or misrepresentation may have arisen which, in a traditional contractual context, would provide grounds for the cessation of contractual performance. If this were to arise in a smart contract context, parties are likely to rely on restitutionary remedies to recover benefits automatically transferred by the code under the smart contract.
A smart contract rendered void on grounds of mistake is likely to provide a foundation for a claim in unjust enrichment, leading to the remedy of restitution. A smart contract rendered voidable on grounds of misrepresentation may be set aside so long as the parties can be restored to their pre-contractual positions. Irish and English law does not require exact restoration. Indeed, it may be impossible to exactly restore benefits transferred by code if such transfers are immutably recorded on a distributed ledger. Provided practical justice can be achieved between the parties—for example, by valuing the benefits transferred by the code in monetary terms and ordering one party to make restitution to the other in monetary terms—practical justice may be capable of achievement.
Damages are still an adequate remedy for breach of a smart contract. The automatic nature of a smart contract means that a breach of contract is likely to occur less frequently. Notwithstanding the automatic nature of a smart contract, if a defective piece of code fails to carry out the obligations of the smart contract, the affected party may be able to claim damages.
The remedy of specific performance may be available if damages are an inadequate remedy for breach of a smart contract. A court could, for example, order a party to specifically perform the smart contract according to the parties’ original specifications.
Frustration is likely to be an important issue in a smart contract context. The leading Irish authority on frustration is the Supreme Court decision in Neville & Sons Ltd v Guardian Builders Ltd. The Supreme Court held that frustration of a contract takes place when a supervening event occurs without the default of either party and for which the contract makes no sufficient provision. Such event must so significantly change the nature of the outstanding contractual rights and obligations from what the parties could reasonably have contemplated at the time when the contract was entered into so that a court is satisfied that it would be unjust to hold the parties to the original terms. For example, if there is a system malfunction that prevents performance of a smart contract, frustration may be arguable. In this regard, force majeure clauses will be of particular importance. Parties may wish to include such clauses in the natural language component of their smart contract so as to identify events that will affect performance, the effect such events will have upon their contract, and what is to occur if such events materialise.
Consumers and Smart Contracts
While existing consumer rights protections were not designed with smart contracts in mind, there is no reason why they should not operate in a smart contract context. However, it is worth considering whether or not such protections are currently flexible enough to protect consumers entering into smart contracts. The European Communities (Unfair Terms in Consumer Contracts) Regulations 1995  require standard contract terms to be fair for Irish consumers. A term is unfair if it puts the consumer at an unfair disadvantage or is harmful to the consumer’s interests. Additionally, terms must be drafted in plain, intelligible language. It may be particularly difficult for smart contracts, consisting wholly or partly of code, to satisfy this intelligibility requirement. Traders may need to include a natural language element in their smart contracts with customers that sets out the terms and conditions in full. Directive 2011/83/EU of the European Parliament and of the Council of 25 October 2011 on consumer rights (the “Consumer Rights Directive”) applies to contracts concluded between a consumer and a trader after 13 June 2014. The Consumer Rights Directive gives consumers extra rights when they enter into distance contracts (e.g. buying goods or services online) and provides a 14-day cooling-off period where the consumer can cancel the contract for any reason. This may be problematic in a smart contract context because the code deploying the contract may be automated and it may be difficult to cancel automated performance.
For contracts involving a dispute between EU parties, the Brussels Recast Regulation and the Lugano Convention will apply. Following Brexit, the UK no longer has the benefit of these regimes and Ireland is the only common law jurisdiction in the EU. In April 2020, the UK applied to accede to the Lugano Convention, which governs the enforcement of judgments between EU Member States and countries in the European Free Trade Association. The EU has not yet approved the UK’s application to accede to the Lugano Convention. Moreover, the EU-UK Trade and Cooperation Agreement does not deal with jurisdiction. It is therefore more important than ever to be able to determine in which jurisdiction a contract was concluded, entered into or performed, and whether the above regimes apply in the event of a dispute.
Under the Brussels Recast Regulation, if an agreement does not include a choice-of-law clause, the domicile of the defendant will determine the jurisdiction. However, the pseudonymous nature of some DLT systems may make it more common for parties to enter into smart contracts without knowing the real identity of their counterparty. This will pose difficulties for identifying the proper party and the applicable jurisdiction regime.
The place of formation of a contract is sometimes used to determine the applicable jurisdiction. A contract is usually formed at the moment and in the place where acceptance of an offer is communicated to the offeror by the offeree. In a smart contract context where there may be little or no natural language interaction between the parties or where computer programs interact autonomously, it may be difficult to identify exactly where and when an offer was accepted.
Jurisdiction can also be based on factors connecting a particular legal system to a dispute (e.g. the place of contractual performance or the place where an asset is situated). Again, smart contracts may pose challenges in identifying the geographical location of performance because the obligations under a smart contract may be performed on a distributed ledger rather than at a real-world location.
Smart contracts have the potential to revolutionise the legal services industry in Ireland. The UK Law Commission’s initial research and discussions with stakeholders have identified the following potential benefits and savings associated with the use of smart contracts:
- increased efficiency and lower transaction costs Every participant will have an up-to-date copy of the ledger, access to real-time details of performance of the smart contract, and performance will occur without the need for human intervention. Moreover, smart contracts may provide security superior to traditional contract law and may reduce transaction costs associated with contracting. CommonAccord, an initiative to create global codes of legal transacting by codifying and automating legal documents, including contracts, has argued that inefficient legal document practices are responsible for a very large part of the cost of doing business.
- lower enforcement costs The code executes automatically and therefore the need to take enforcement action for non-performance should be rare.
- reduced risk of fraud The consensus mechanism and immutability of a distributed ledger mean that contracting parties can trust each other and transact in confidence.
While smart contracts can offer many benefits, it is essential that Ireland begins to consider the risks posed by smart contracts and how such risks can be resolved in an innovative way. In our previous article discussing the UKJT’s Legal Statement, we concluded by stating that a number of options presented themselves to Ireland, in particular using the Legal Statement as a kick-off point for achieving a degree of legal certainty under Irish law. We proposed that a consultation exercise, similar to that of the Legal Statement, could be performed in Ireland which could then be presented to the Law Reform Commission for the purposes of review from the standpoint of Irish law and to consider whether any Irish legislation, EU Directives or Regulations might be desirable in the area of smart legal contracts. Unfortunately, one year has passed and Ireland has still not reviewed the current legal framework to ensure that Irish law facilitates the use of smart contracts.
The UK is making strides to ensure that the use of smart contracts is regulated in its jurisdictions. This can be seen by the UKJT’s publication of the Legal Statement in November 2019 and the Law Commission’s Call for Evidence in December 2020. While the Legal Statement has no legal standing, it will likely be relied upon by the Law Commission as part of its scoping study requested by the UK Government. The Law Commission is currently carrying out the Call for Evidence and will publish the results of the scoping study in late 2021. The Law Commission is also intending to launch a consultation paper on digital assets in early 2021. Clearly, the UK can see the economic risks of failing to adapt the law to digital change.
The UK is taking the first steps to identify the legal issues regarding smart contracts and is attempting to find solutions. This is commendable. To remain a competitive choice for legal services and the resolution of disputes post-Brexit, Ireland must now begin reviewing its current legal framework to ensure that it facilitates the use of smart contracts. As identified by the Law Commission, there are questions about the circumstances in which a smart contract will be legally binding, how smart contracts are to be interpreted, how vitiating factors such as mistake can apply to smart contracts, and the remedies available where a smart contract does not perform as intended. It is time for Ireland to begin examining the circumstances in which a smart contract will be legally binding, how smart contracts are to be interpreted, how vitiating factors such as mistake can apply to smart contracts, and the remedies available where a smart contract does not perform as intended.
Pursuant to s.4(2)(c) of the Law Reform Commission Act 1975, the Attorney General may request the Law Reform Commission to exam specific areas of law. To avoid being left behind, the Attorney General should request the Law Reform Commission to commence a scoping study analysing current Irish law as it applies to smart contracts, identify areas where further work or reform may be required, and provide options for reform. The nascent state of the technology used in smart contracts means there are few, if any, clear answers on the legal issues presented by smart contracts. But in uncertainty there lies opportunity, and this opportunity is one that should not be missed by Ireland as the only common law jurisdiction in the EU. It is now time for Ireland’s lawyers and technologists to come together and answer Ireland’s call for evidence which, it is hoped, will be sooner rather than later.
The author wishes to thank Colin Grant and Grainne Bennett for their contributions to this article.