Thoughts on formalities and artificial intelligence

Sat 10 November 2018, Philippe Ropenga Artificial intelligence

Recent blog posts have dealt with will drafting¹ and formalities². They are the substrate of general thoughts on formalities and the legal profession during the rise of artificial intelligence. Lawyers have to take care of formalities for their clients. It therefore is interesting to examine the link between law and formalities to see how artificial intelligence comes into play (1). The rise of artificial intelligence encourages the lawyer to ask himself whether he is a manufacturer or a craftsman (2).

1. The link between law and formalities

Formalities in general are used to protect the precious and the weak (1.1). When lawyers check facts and apply legal rules, they tend to assimilate facts of a given matter to a well-known situation (1.2).

1.1 Protecting the precious and the weak

Land registers are a mean to ascertain rights and to make sure that anyone interested is aware of the status of a given piece of land. Formalities may also protect weak parties by imposing a heavier burden of proof on one party to favour the other one. Mainly after World War II law has focused on identifying weak parties. The archetype of the weak party is the consumer who is more or less informed and more or less clever, presumably less than more, according to current trends regarding consumer protection. This consumer-like approach of any weak party tends to put the spotlight on information rather than on global understanding of a given situation by a weak party. Will drafting in the context of incapacity is an example of such a change. The purpose of this post is not to discuss the issues raised by the test set out in Banks v Goodfellow³. A recent consultation paper issued by the Law Commission mentions that it is outdated⁴ and compares it to the test set out in the Mental Capacity Act 2005⁵. Looking at this question from another country allows to see that section 2 of the Act focuses on the processing of relevant information by the incapacitated person. Understanding is thus seen as related to data processing. This certainly has something to do with the better understanding of cognition by neurosciences and medicine. Neurosciences have in turn influenced the development of artificial intelligence. These issues are really interesting but are beyond the scope of the legal practitioner's daily practice. It however is important to note that when the practitioner pays a greater attention to information, he paves the way to artificial intelligence since information technology can effectively handle information to perform standardised tasks whenever standardisation is based on processing information.

Let us now see how a practitioner processes information in practice.

1.2 Dealing with well-known situations

The practitioner analyses facts to determine what piece of legislation is applicable. He then analyses them to apply the law and suggest a solution to the client. Artificial intelligence is already largely used at the first stage of the process. Legal databases reduce the time that is needed to find and grasp relevant materials. It also is largely used at the final stage to reproduce at least partly standardised documents. The standardisation of formalities is a safety net. A practitioner wishes to apply solutions that lead to a predictable result. A given matter thus tends to boil down to a well-known legal situation. This explains the success of smart contracts. Furthermore, securing a transfer of legal title may be easier with blockchain. Nevertheless, the practitioner who deals with successions, family estates, or family businesses knows well that the way in which a client understands the situation that he faces largely depends on his feelings about this situation. One could say that feelings are biases that have an effect on the result of the processing of information by the client. The practitioner deals with emotional issues on a daily basis. The effect of emotions on understanding is beyond the reach of artificial intelligence today. It means in practice that if the practitioner has to rely on formalities to protect, for instance, a legal title or a vulnerable person, he may benefit from artificial intelligence but cannot rely only on formalities and procedure to consolidate the weak points of a case since these points often have little to do with procedure or formalities.

It has been seen that artificial intelligence really was interesting in practice. It also seems that information technology applied to law requires the practitioner to reflect upon his practice.

2. The lawyer: manufacturer or craftsman?

The practitioner has to determine whether he conceives legal assistance as a delivery of legal services that will be consumed by the client (2.1) or as a mean to develop value (2.2).

2.1 The consumption of legal services

The effect of artificial intelligence on legal practice is not an easy question. The practitioner may know what blockchain is and to what extend smart contracts can be useful but still feel that he will be entangled in expensive subscriptions if he makes any move to add artificial intelligence to his work. There are two reasons to this reaction: practitioners who have a monopoly to do certain things may feel less professional if they delegate these tasks to artificial intelligence, and they may not know exactly how to coordinate artificial intelligence with human skills. It has been seen in a previous post that many disputes arise because of issues related to the allocation of time, money, or other resources.⁶ Distributions decided by a family member and related disputes are not purely rational and often involve emotions. Drafting sophisticated documents with smart contracts may be quick but not necessarily useful to solve difficulties that have an emotional background.⁷ A practitioner in the field of family estates or family business may therefore be reluctant to subscribe to a service that gives him partial information, whereas he accepts to pay for a database that allows him to find an answer to a purely technical question. In other words, artificial intelligence can be useful when one has to reproduce similar documents since it can help the lawyer to customise them effortlessly and to reduce the risk of mistakes regarding formalities. Artificial intelligence would have a similar function to robots in factories. A practitioner who deals with complex issues that are not only technical is presently less likely to benefit from artificial intelligence. It however does not mean that a practitioner who mainly solves his clients' issues by creating bespoke solutions would not benefit from it.

Let us now see how a practitioner who solves issues without drafting any deed could benefit from artificial intelligence.

2.2 The development of value

Crafting custom solutions requires to determine what element may be used as a starting point to create a new situation. It can be a statutory provision, an international regulation, a fact, the attitude of a party, or a combination of several of these elements. The practitioner main task may, for instance, not consist in drafting a contract but determining what a foreign legislation says on a specific issue to then determine if triggering the application of a foreign law is useful. The practitioner thus has to rely on databases to search for relevant information. Publishers tend to provide more and more content. They also are working actively on blockchain and smart contracts. Databases and specialised search engines have not made any substantial progress while vocal assistants that involve speech recognition as well as main search engines are much more efficient now than they used to be ten years ago. Tools that are used by the general public should be more difficult to conceive than specialised search engines since they have to take a variety of contexts into account. It seems to me that legal publishers tend to abandon databases in the sense that no significant improvement regarding search engines has been made so far. Smart contracts and blockchain may seem more appealing than databases. I however regret that publishers focus on providing more information instead of providing more relevant information.

It has been seen that implementation of artificial intelligence was a complex issue that had to do with the orientation of the legal practice and with the strategy of a publisher.

In brief, artificial intelligence will affect any legal practice but unfortunately legal practitioners do not seem to be leading this change.

Go Top