Artificial intelligence-driven robots are deployed in all echelons of companies to perform repetitive tasks. But what if you had a robot sit on the board of a joint stock company? Is that possible under company law? And is it at all desirable? These two questions were central to Adam Nassiri’s thesis. We spoke with him about his motivations and findings.
“I had little knowledge of artificial intelligence (AI) at the outset,” Adam Nassiri confesses. He is nonetheless aware that robots are bound to play an increasingly important role in society and in companies. His interest was consequently aroused when he read an article about AI in corporate law. “I wanted a topical and interesting subject for my thesis and that article got me thinking. I also wanted to make a contribution to jurisprudence.”
Equating robots with humans
“There could certainly be financial advantages if a robot were to be appointed on the board, for example, because you want more leeway as to the country where you want to settle,” Adam Nassiri says. The question boils down to whether the advantages outweigh the disadvantages. He therefore decided to investigate whether a robot as a director of a joint stock company is legally feasible and desirable.
He cites the robot Sophia, which is registered as a citizen, by way of example. “I find that quite extreme, because you then equate robots with humans, whereas robots do not yet have human qualities such as spirituality, emotions and feelings. But creating a legal entity for a robot makes more sense, because a legal entity is created to serve humans.”
Robots serve humans
“Other legal entities include e.g. a private limited liability company, municipal and other government agencies, etc., all of which have legal personality. They too serve humans. So if a robot is defined as an entity that is demonstrably there to make the life of humans easier, then it would be possible to appoint it as a director under Dutch company law. Only one question then remains: is it desirable?”
There is no unequivocal answer as to whether “it is desirable for a robot to be appointed as a director.” Advantages can certainly be cited, such as increasing the freedom of establishment, but also speeding up decision-making processes. In China, they have already appointed a robot as a director for this very reason, and several foreign jurisdictions are considering and have already adopted regulations on robotics and artificial intelligence to some degree.
A robot falls short when it comes to charting strategy
“From the perspective of legal theory, I would say that it should be possible to create legal personality for robots and by extension a robot director,” says Nassiri. But it becomes more complex if a robotic board member has to take part in the decision-making process when charting strategy, for instance. A robot will then fall short if it is designed to analyse financial statements, for example. That will in turn breach the principle of collective management and responsibility.”
Suppose a robot is appointed as a director after all. What happens if it then makes wrong decisions? “The European Parliament has tabled a proposal that companies should take out compulsory insurance to cover a robot director that is held liable. But supposing that an insurer will not or cannot pay out, then you can set up a general fund for robots from which such payment can be made.”
The robot as a digital person
If neither the insurance nor fund can provide relief, then there is a third option, because the robot director is considered a digital person and “with the amendment of Section 2.11 of the Civil Code, whereby the digital person and the manager are included in the relevant article of law, damages can be recovered from the manager of the robot director in question.”
In any event, as a joint stock company you must have enough trust in the robot, which ostensibly gives rise to an ethical issue: is it responsible to use a robot director?” It is certainly an ethical matter because you then run more risks: Cyberattacks, for instance, but also discrimination, programmed by the robot’s developer, whether consciously or unconsciously.”
Not yet a director
Nassiri therefore does not see the robot as a director of a joint stock company as being viable, but AI experts in law beg to differ. “I read in one article that it will take another sixty to seventy years at the earliest, but in another article that it would be possible in the foreseeable future. More and more voices are being heard that robots should be assigned a legal personality, but that is still minimal for directors.”