Blockchain, Smart Contracts and AI in the Legal Market13 min read

In 1995, the journalist Clifford Stoll made the following statement, “The truth is, no online database will replace your daily newspaper, no CD-ROM can take the place of a competent teacher and no computer network will change the way government works. […] Nicholas Negroponte, director of the MIT Media Lab predicts that we´ll soon buy books and newspapers straight over the Internet. Uh, sure.”

The digitalisation era is the biggest transformation since the industrialisation, and it enables us to do previously impossible things as well as rewrite the standards on how things can be done. The opportunity to see and analyse patterns from general and individualised data gives us a deeper knowledge of the client’s individual needs and requests, which leads to better services with higher user quality. The general demand today is to make a product accessible in a short amount of time, at any time, whether we are talking about music, food or legal services. The client wants their lawyer to be accessible when they require it. If we in the legal industry do not meet this demand and challenges the way we deliver legal services, we will get stuck in a conservative and slow way of doing things, justified with continuity, and eventually become out of date.

However, the existence of an extreme amount of information and individualised data challenge the technical systems and the legislation we have today. How we can and should use this information is an even bigger challenge in itself. Therefore, it is of highest importance that there are secure and transparent digital techniques for transactions and storage of information that can meet the requirements from both clients and the industry – not just from a legal perspective, but also to match the digital and connected lifestyle people have. Businesses on the other hand, cannot only keep up with it, but must be in the very forefront of a changing society and need to create new services that can meet the client’s demands.


The blockchain revolution is still in its development process. Blockchain was introduced in 2008 with a white paper by Satoshi Nakamoto with the purpose to serve as the public transaction ledger of the cryptocurrency Bitcoin. However, blockchain can be used for so much more – especially in the legal market. A simplified explanation of what blockchain is could be described as a vault for storage of information, where data already added to the vault can never be changed or modified without notifying the others having access to the vault.

Blockchain involves cryptographic hash functions, which are comparable to a digital fingerprint. The way this works is such that anything that is storable as a digital file (i.e. pictures, agreements, movies, transactions, documents etc.) is assigned a unique code – a fingerprint or hash. The algorithm used to create the hash, gives every digital file a unique and unpredictable code, equivalent to a fingerprint but in numbers. Every so often, transactions are bundled together, and their hash functions are grouped in blocks – or vaults, using the previous analogy. The transactions can be controlled and authorised separately but are grouped in blocks to enable the spread of multiple verifications at once. The authorisation of the transactions consists of checking that the transaction has been verified with the owner’s password and that the owner has not tried to execute the same transactions multiple times. When the block is authorised, it is linked together in a chain of blocks, where the following block include a verification from the previous block. Each occurrence like this is documented in a verification list. The list of verifications is then distributed to the ones having access to the blockchain.


When information is stored in a blockchain it is not stored at one place, like with banks or other centralised systems but shared with the thousands of nodes in the network. The networks constitute a peer-to-peer network, or a distributed ledger system that is decentralised and immutable. Since every new block brings a verification of the previous block with them, it is impossible to add information in an already existing block without changing the entire chain. Blockchains are intended to be permanent, tamper-proof records that sit outside the control of any governing authority. This is what makes it such an attractive and useful technology; there is no single point of failure, i.e. no component or person that can put the network at risk. There is no eventuality of human errors and it is immune to corruption or blackmailing.

With help of blockchain, parties can register and verify transactions, documents and agreements with full insight and trust, despite the absence of a trusted third party, such as a bank or an authority, which normally would be a middleman in a chain of transactions with the purpose of validating the accuracy of a transaction or an agreement. Trust has always had a fundamental function when it comes to transactions, which is why we have created a world of trusted third parties, like banks or enforcement authorities. With the use of blockchain, parties do not have to trust or even know the opponent because of the possibility to verify the transaction or document themselves. Therefore, the party’s credence of the opponent can be completely non-existing, but still have full control over the transaction and its execution without the involvement of a trusted third party. This makes blockchain transactions more secure, cost efficient and transparent than other techniques used today.


Blockchains can be divided into separate groups, in which the main distinguisher would be between open and closed blockchains. Blockchains can also be partly open and partly closed. In a private blockchain, it is only one or a limited number of actors that can authorise the hash functions in the blockchain. For example, this can be a company’s internal database or an authority storing personal information, only accessible by the affected person with a private key. In an open blockchain, accessible to everyone, practically anyone can authorise the blocks, according to already set rules. However, according to GDPR, it is not possible to store personal data in an open blockchain since it is not, using GDPR terms, erasable (“the right to be forgotten”). Nonetheless, personal data can be stored “off-chain” and the reference to this data stored along with a hash of this data in the blockchain.

Another way to use blockchain in a GDPR-proof way is to anonymise the data. However, the anonymisation is not always bulletproof, which makes this the least preferred solution of the two. A third alternative is to make a “centralised decentralised system” where authorities own the blockchain, and the records are made official documents according to the principle of public access to information. At the same time, third party service providers can be let into the system without the information being considered official through a practice called “own space”, which is already used for e-services. That type of information is considered to be under technical processing and not received by the authority, being therefore excluded from the principle of public access to information.


The digital revolution has made a non-paper-based system standard for carriers of information, which shows that the biggest transformation has already happened. Lantmäteriet, an authority belonging to the Swedish Ministry of Industry, launched their blockchain-based service for property transactions in 2018. However, this is just the beginning of what blockchain can be used for and how it can change legal services. Data protocols based on blockchain, like smart contracts, with a purpose to verify, negotiate or compose agreements, leave no place for human errors by automatically executing transactions when a predefined condition is met. Smart contracts are characterised by executing themselves, without any exceptions, which creates full trust for both parties. An example of a smart contract could be the following: A lawyer has been asked to create a will where estate A is allotted to person A and painting B is allotted to person B. The lawyer draws the will and enters the parameters into a data protocol – a smart contract – which is constantly updating itself towards the Tax Agency’s register to see if the testator is still alive. When the testator dies, the data protocol will know and immediately execute the will. For the sake of the example, the estate and the painting has already been liquidated, and the money can therefore be transferred to legatee A and B the same day the Tax Agency is notified of the testator’s death, all without any human input.

The same technique can be used in a commercial transaction where a payment will not be executed until a product is delivered. There could also be a predefined condition stating that in case of a breach, a specific amount will automatically be transferred to the other party. A smart contract can therefore be used as an escrow agreement. The point is that, with smart contracts, both parties must follow through until completed, which creates higher transaction security than normal transactions. A desired effect is also that unclarities of wording in normal agreements that can lead to a dispute will disappear. With a smart contract, there should not be any misunderstandings about intent or interpretation due to the already set parameters and automatic execution of the transaction. Adding AI to a smart contract further increases its efficiency with reducing the need for human verification, intervention or analysis. The combination of AI and blockchain based smart contracts simplifies the negotiation and execution processes, which will ultimately lead to greater efficiencies.


In a vendor relation, a built-in computer in a fridge or storage closet can register the usage of groceries and store the information in the blockchain as AI. The same computer would then be able, with this information, to create a smart contract and purchase groceries with deliverance to the door. This is with no doubt possible, but the question is if the transaction is legal, according to existing contractual principles and rules. If it is legal, there is still a question about liability that must be answered. With the use of AI, software can be given the possibility to steer away from the original contract by experiencing, draw conclusions and learn from its mistakes and in that way program itself. That the software would start programming itself out of control of the involved party is highly unlikely to ever come about, but we have not yet seen the full potential of what AI can be used for in contractual matters, and what challenges that may include. In a digital world, we trust the software and its ability to make decisions which can lead to a constant and uncritical influence of the ignorant mass. Therefore, it is important that there are legal models for liability of transactions made with limited or no human input.

Error in transactions made with smart contracts should be solved with liability for the party with the best possibility to be in control of the transaction, i.e. the supplier. This would involve liability to indemnify if the supplier does not prove that the failure was due to an impediment beyond their control and that they could not reasonably be expected to have taken the impediment into account at the time of the conclusion of the contract, or to have avoided or overcome it or its consequences. This is applicable to error in transactions, however, it is not as clear to ascribe liability when the error lies in the collection and use of information underlying a smart contract. The legal position on liability on error in this situation must be defined as unclear. However, in such situations, the party deemed responsible, should be a  party with the ability to control that the collected information is used in a predestined way and that the construing of information is proximate. Depending on the matter and the construction of the smart contract, the liable party could on one hand be the software developer, or on the other hand fall back onto the supplier if the developer acts on the behalf of the supplier or as stated in an agreement between the developer and the supplier. Nonetheless, it is possible that there is room for a partial disclaimer.


When the Swedish Contracts Act was formed, there was no possibility to anticipate all the applicable areas that the legislation covers today. Parties have the freedom to arrange their conclusions of agreement, but some elementary demands must be fulfilled for the contract to become legal. The Act is tech-neutral, and an electronic agreement between two parties usually does not introduce further complications. However, with agreements through digital elements it can be difficult to distinguish two equivalent declarations of intent, which by the main rule is a necessary requirement for the conclusion of an agreement. When a smart contract is made automatically, with limited or no human input, it is difficult to discern a party’s expressed intent. Therefore, when it comes to smart contracts, the Act must be construed, which can cause uncertainties if the legislation is applicable to contracts made with no or limited human input.

A computer is not a legal entity and cannot undertake any legal acts. Therefore, two perspectives have risen in the legal doctrine to explain how automatic transactions still can conclude a legally binding agreement. Since a computer lacks legal competence and lacks the ability to express intent, the argument refers to the agreement between parties as a hypothetical intent that is being reflected in all future deals. By “signing up”, the parties agree on an automatic routine that includes future agreements. However, since the parties lack knowledge about the specifics of the future agreements, it could be argued that an actual declaration of intent is missing.

The other argument is a more objective view on the closing mechanism of an agreement, and points at the surrounding circumstances of the agreement. The effects of the agreement occur as a result of the combined circumstances, and therefore the agreement becomes binding. In this case, the traditional declaration of intent does not exist, but the effects of the agreement arise through the specific situation. Intercessors of this standpoint claims an argument for a hypothetic intent to be forced and constrained, in comparison to an objective view of the express of intent that is ductile to a complex system. According to principles within the Swedish contract law, a person presenting intent can fulfil the requirements of declared intent, whether a conventional way of communicating this intent is used or not. In the case where a visible circumstance that impresses a declaration of intent exists for the other party, the requirement of declared intent is met.


In summary, it is not clear what principles of Swedish contract law that apply when it comes to smart contracts concluded through blockchain-based systems, with no or limited human input. Despite that, there is an endorsement that the parties’ intents have been declared and fulfilled. The Swedish legislation does not in itself constitute a holdback for the use of blockchain or smart contracts. The legislation applicable in normal situations concerning conclusions of agreement, should in large parts be directly applicable to contracts made with limited or no human input. Over time, case law will be developed and models for interpretation will help to clarify the implementation of blockchain and smart contracts.

Jennie Gunnarsson