Which legal practices could be automated?

“Do we want artificial intelligence to speak the law?” - IT and media lawyer Björn Steinrötter researches legal issues relating to digitization

He stands in front of students in the lecture hall, is the author of legal essays, commentaries and manuals and researches the latest legal problems relating to artificial intelligence, Bitcoin and the data economy. It helps to shape areas of law that do not yet exist. Björn Steinrötter is junior professor for IT and media law at the University of Potsdam. He is one of twelve young researchers who were able to secure one of the coveted tenure-track professorships that the University of Potsdam announced in 2018 with funds from the federal-state program. More than 1200 applications were received for the four open-topic professorships alone. His career could have been very different.

Björn Steinrötter is a trained banker. “My father advised me: 'Do' an apprenticeship first. Then you have something in hand if your studies don't work out ‘”, he says and laughs. The concern was unfounded. With his training in his pocket, he decided to follow his dream and began to study law in Bielefeld. From the first day of his studies he was enthusiastic, says Steinrötter. “In retrospect, it was a great time, one of the best of my life.” But he not only liked the law, but also science early on. During his studies he worked as an assistant at Professor Staudinger's chair in civil law. The young graduate was less interested in legal practice, as he discovered during his legal clerkship. “I originally wanted to be a lawyer. But I don't seem to be particularly keen to represent someone whose position I totally disagree with. So I turned to research all the more enthusiastically. "

No "Jura air castles"

After completing his doctorate in private international law in Bielefeld and completing his legal clerkship in Berlin and Sydney, the civil lawyer started as a postdoc at the Institute for Legal Informatics at Leibniz University in Hanover in 2015. There he worked his way up to IT law under Professor Heinze - a decision he has not regretted. A certain basic understanding of computer science is of course essential for this. “But you don't have to be a programmer either. The exchange with IT specialists, engineers and practitioners is important and helpful to me, ”says the IT lawyer. “The feedback I get on this path ensures that I don't drift off into 'Juraluftschlösser'.” So even hotly debated theoretical problems sometimes turned out to be rather unrealistic. Like the almost classic model case from the field of autonomous driving that fills entire conferences: suddenly several people run into a self-driving car and the decision is made to run over and injure either an older couple or a young mother with a child , maybe even kill. Can this dilemma be resolved ethically and then legally? “An engineer who looks at it, of course, approaches these questions differently. Because apart from the very low probability that such a case will occur, the sensors and algorithms may not be able to reliably differentiate the mother with the child from an older couple, nor can they implement ethical guidelines one-to-one. Because, unlike the deterministic ethical model of thought, for example, they are not able to determine whether injuries or death are certain to occur. Reality and thus the programming based on probabilities have to deal with such uncertainties as a central accident characteristic. It is therefore more in the direction of 'risk ethics'. "

IT law as a cross-cutting issue

Björn Steinrötter crosses borders with his area of ​​expertise, but also in law itself. “IT law is what you would call a cross-sectional matter. While traditionally the classification is mostly based on the legal field, in IT law it is the object of consideration that characterizes the discipline - in principle from all conceivable fields of law. I cannot therefore cover the entire spectrum; I focus on the civil, data, intellectual property and conflict of laws perspective. "

Recently, he has dealt intensively with problems relating to the booming world of crypowers such as Bitcoin & Co, whose legal security has so far barely kept pace with technical developments. “The trading of such virtual currencies works across borders and on a decentralized basis. The question therefore arises as to which law should apply to the individual transactions, ”says the researcher. In the context of commercial law, the principle that the law of the market place is to be applied has so far often applied. But that has now disappeared, where business is carried out directly and decentrally without geographically localizable financial markets. "The general regulation, which you would then regularly revert to, now states that the law of the country from which the seller comes to apply," says the lawyer. "But since more and more such transactions take place pseudonymously or even anonymously, a new, factual problem arises."

Another specialist area of ​​the IT lawyer concerns the legal issues surrounding the increasing presence of robots and artificial intelligence (AI) in our everyday lives. Steinrötter is Deputy Chairman of the Robotics & AI Law Society (RAILS), which brings together representatives from various disciplines precisely for this purpose. Among other things, he researches legal problems in geriatrics. “There are currently many questions about the use of robots in care, especially in old age: Who is liable for technical defects and accidents? What about data protection? And what are the requirements in terms of IT security? ”Asks the lawyer. A manual is to be published in spring 2020, which will discuss the most important current legal issues relating to AI and robotics on around 1200 pages. Steinrötter is one of the editors. The particular challenge here is - as with IT law in general - to see to what extent the existing rules can adequately deal with the new questions. "Sometimes it seems as if the applicable laws no longer apply and new ones have to be created," says the researcher. However, on closer inspection, they often turn out to be forward-looking and well-formulated - and may only have to be interpreted accordingly or, if need be, specifically adapted. “In any case, with the manual we are already looking far into the future and sometimes discussing cases that do not even exist yet. That rarely happens in law. "

“Legal tech” could change legal practice

The legal industry itself is currently experiencing a real digital boom - due to the emergence of so-called "legal tech" applications. This includes various forms of digitization: On the one hand, more and more smart tools are coming onto the market that are intended to make work in law firms and legal departments faster and more efficient. For example, you can automatically create simple pleadings. But law enforcement itself could also be made faster and more efficient with their help. “That is to be welcomed in the starting point. At the same time, of course, the question arises: How far should this go? ”Says the researcher. “Do we want Artificial Intelligence to handle all of our legal processes and, in the end, to speak the law itself at some point?” These are discussions that have been part of the legal informatics discipline since the 1950s, but which have largely disappeared from the university landscape after a few fruitful decades be. Now it was primarily legal practitioners who promoted similar debates, which are now known as 'Legal Tech'. “These topics also belong to a university. You can't hide from it, no matter how you feel about it. ”For this reason, Björn Steinrötter has already organized a two-day workshop on“ Legal Tech ”for the summer semester 2020 - for computer scientists and lawyers alike. Together, the participants should first work out the technical and legal basis for such a tool, then go to a law firm and see whether it can solve real cases. “Simple and trivial cases can certainly be processed more quickly with good software. But I am quite skeptical of the current hype. Our legal system does not follow a formal logic, but a logic of values. It is not always about the binary decision right and wrong, but about justifiable and not justifiable. ”He could therefore not imagine an automated system of jurisdiction. “But I have the feeling that a lot of young people would prefer to be tried by a machine - because they hope to be treated so without prejudice. But this is a mistake. Algorithms and especially their training data can also have prejudices. But: Then we not only have one or the other 'crooked' judge, but a whole 'crooked' system. "

The tourism industry has been experiencing a similarly rapid leap into digitization for some time - and here too Björn Steinrötter is involved in “reclaiming” new legal territory. He recently wrote up-to-date comments on the EU Air Passenger Rights Regulation and travel contract law, and is also co-editor of the trade journal for tourism law “ReiseRecht aktuell”.

The fact that Steinrötter's specialty is still anything but the well-trodden path can be seen not least from the fact that it can only be taken optionally in law studies. He teaches at the Law Faculty in Potsdam, especially in the area of ​​"media and business law" as well as in the master's degree in media law at the Erich Pommer Institute, in which the University of Potsdam is involved. But that should change - also within the classic disciplines. “Digitization will have to deal with more and more areas of law. Also, for example, labor, company or administrative law. And in addition to the data protection law strengthened by the General Data Protection Regulation, a data management law is increasingly emerging. After all, data has become an extremely important commodity. "

Prof. Dr. iur. Björn Steinrötter, Faculty of Law
Email: steinroetteruni-potsdamde
Internet: https://www.uni-potsdam.de/de/lssteinroetter/


This text was published in the university magazine Portal Wissen - Zwei 2020 “Gesundheit”.