Can Artificial Intelligence Replace Legal Expert Knowledge?

Artificial intelligence (AI) has already displaced many employees from factory floors and has gained a firm position in banking and finance. The first fully automated helpers are also making themselves useful in hospitals and on strawberry plantations. Law firms and legal departments, however, are still dominated by the human mind. Lawyers always have to think through new problems that arise from the individual case, because at no point in life is one situation exactly like another. This is why the opinion persists in legal circles that solving legal problems requires customised work that a machine can never do.

Recently, a large law firm in the USA introduced the first lawyer robot, which, when approached, conducts research on the desired case and compiles materials with references. So-called legal chatbots, which provide information in response to a legal question entered, are already used in the USA as an alternative to costly legal advice. But will human lawyers soon be completely replaced by robots? Could the robo-lawyer soon don the robe and appear in the courtroom? No, it is not possible to envisage such scenarios in the foreseeable future. Rather, in the coming decades, the use of legal tech will remain limited to narrowly defined areas and will only include routine work.


Proven approaches to AI-based systems

In traditional software development, a programmer creates a rule-based system, which means that they must define in advance all the rules according to which the software is to function. AI is a generic term for all those systems in which the programme learns itself, i.e. can also surprise the programmer. These non-rule-based applications include, for example, natural language processing, machine learning and predictive analytics. In Natural Language Processing, the software reads a text written by a human being and understands its content, as far as this is possible for it. In any case, the program can grasp the context from one formulation and likewise extract these statements from another formulation. So instead of entering search words, users can describe their problem in their own words and get an answer from the system. Another working method of AI that has already been tested today is the evaluation of large amounts of data in order to make predictions about the probability of future events (predictive analytics). Finally, machine learning allows the system to learn from trials and become more accurate in solving certain tasks over time. Most legal operations today are based on one or a combination of these systems.


Legal tech in contract drafting and review

In everyday legal work, AI-based systems can help above all in the drafting and review of contracts. For example, sales contracts for real estate only differ from each other in terms of personal data and individual clauses, while around 15 out of 20 text modules are identical. If the data of all parties involved and the property are known, an intelligent software program can create the complete contract in seconds. In doing so, an agile and already experienced software knows which formulations are legally sound and which clauses the contract should not contain.

The AI can also quickly find the relevant passages when reviewing contracts with regard to certain formulations that are no longer permissible and save the lawyers hours of file leafing. The application also points out unusual omissions or other deviations from the standard. Some programs for contract review have already proven to be very efficient and cost-saving in practice, for example an AI-based application from Legal Analytics GmbH.

The legal profession can also benefit from a prognosis function to a limited extent, because there is already software that makes predictions about the outcome of court proceedings in a particular court. Based on the collected decisions of the judges involved, the program not only calculates a probability of winning or losing, but can also recommend which argumentation is most likely to be convincing. And finally, a machine can at most learn to think in a rudimentary way, but it can never make evaluative decisions. At best, it is able to list arguments pro and con and thus prepare the decision-making of lawyers.

AI-based systems can be of great use in drafting and monitoring contracts by increasing output, reducing errors and speeding up work processes. They can also be used for the tedious research that is currently mostly done by junior colleagues in large law firms. Machine responses to simple legal questions can also facilitate and speed up workflows within a company or government agency. However, these innovations do not make human lawyers superfluous, because they are needed as controllers and have to put the finishing touches on any automatically generated work. For the rest, they can spend their saved time on the tricky questions or devote themselves to other meaningful activities. Perhaps the young lawyers in large law firms, who today spend their working day with tedious searches for references, will be able to look forward to more interesting activities in the future.


Possibilities and limits of AI in the legal field

Artificial intelligence cannot currently replace human intelligence completely and probably will not be able to do so in the future. Although software is many times superior to the human brain in certain areas, these are insular talents. After all, human intelligence consists not only of IQ, but also includes important social and emotional skills. And legal work is characterised only to a small extent by such routine tasks that a robot could take over. For tasks that leave the narrowly defined field, machine helpers quickly reach their limits. This is because they lack any background knowledge and a sense of context. Intelligent software can recognise a pattern from pictures or words, so that it can distinguish the image of an orange from that of a banana, for example. But it does not know what fruits are, where they grow, how they taste or what significance they have for mankind. Lawyers deal with texts that are not always easy to grasp even by a human mind without previous legal training. The interpretation of legal concepts and legal texts is only successful if the context is taken into account, which always requires historical as well as cultural background understanding.

Share this article

Share on facebook
Share on twitter
Share on linkedin
Share on email

More articles


NDAs Usual Duration

NDA Usual Duration If in the course of the business one needs to transfer information that requires confidentiality, a non-disclosure agreement is a proper tool

Read More »

Leave a Reply

Your email address will not be published. Required fields are marked *