Law 

From robotisation to artificial intelligence. In legal services

Like in other industries, innovation will gradually set the new rules of the game in law. In the near future, clients will stop paying lawyers entirely for certain services. Or they will be willing to pay only negligible amounts. However, there is already a whole range of technologies that can move the efficiency of legal services forward, from basic “robotisation” of legal advisory to the involvement of artificial intelligence. Where on this scale is the right solution for you?

Some people look at the current technological development as a transition to a massive use of automated document preparation. Others see this process as a long wait for the right artificial intelligence that will change the whole industry. Both views are good. But both are also wrong.

Transformation under the influence of technology

The legal field is so varied that the journey of its transformation under the influence of technology will also have to be varied. Above all, there is no one correct “automation” for the entire legal industry. There is a whole scale of possible solutions.

The simplest degree of development is robotic process automation (RPA), where robotic software essentially simulates human behaviour consisting in simple repeated tasks such as entering data in documents. The second degree is cognitive automation, where machine learning and advanced analytics are used to imitate human judgement in simple situations. Here, we can imagine text analysis, fraud detection, intelligent FAQ communication with the client etc. The third degree is artificial intelligence, which should essentially expand human intelligence. At present, it is represented by the interpretation of complex contracts including the option of their approval in the system if they do not deviate from the standard. Prospectively, it will also include complex interactions with the client.

Three steps to victory
An example of a first degree solution is a system which is able to generate a new set of documents when a legislative change occurs, is identified and entered into the system, for example by the in-house legal department, and then to send the documents by email to the client for acceptance.

An illustration of cognitive automation is, for example, the GDPR chatbot that can be asked questions in normal language. The chatbot interprets the question, then asks several additional questions to narrow down the problem. If the resulting question meets the required criteria, the chatbot provides the answer itself. However, if the question involves a risk exceeding a certain threshold, the communication with the client is redirected to a human.

Finally, an example of a solution including elements of artificial intelligence would be an analytic tool able to go through a large portfolio of documents as part of an acquisition due diligence review, e.g. documents related to syndicated loans with the law applicable to documentation in several jurisdictions. The assessment tool applies a set of basic questions to each document. Thanks to this it subsequently identifies documents that the transaction team should focus on – in other words, documents requiring human review. The tool significantly increases the efficiency of the time spent by specialists on extensive transactions.

However, what is most important is the fact that these examples are not just a theoretical possibility. They are something that the market already offers to legal service providers and in-house legal departments.

Automation is an opportunity
It is key to use innovation where it can save costs while also mitigating risks. It is therefore probable that at present, many clients and in-house departments will have use mostly for a form of automated document preparation. There is effectively no reason why automated document production should not be introduced for a variety of processes. On the contrary, it usually decreases the risk arising from possible human error during the creation of a large amount of repetitive documents.

But this is certainly not the end. It is noteworthy that artificial intelligence able to beat a human at poker now exists. Poker has been called a game with incomplete information, and as such it is very close, for example, to contract negotiation. What we can be pleased about is the fact that this AI was developed in cooperation between Czech and Canadian universities.

So when will the time come when the colleague sitting across the table from us at a meeting will look at their tablet all the time, where an AI will give them advice on the best course of action? The technical solution that would allow this is already available.

The transformation of the legal industry has therefore already begun. What first step on the scale of automation options will you choose, today?

An abbreviated version of this article was published in the supplement to the Ekonom weekly dedicated to Innovative Lawyers 2018, 7 June 2018.

RPA

Upcoming events

Seminars, webcasts, business breakfasts and other events organized by Deloitte.

    Show morearrow-right