“Keeping pace with technology” – A conversation on the future of law with Prof. Ugo Pagallo (part 3)
Ugo Pagallo is Professor of Jurisprudence at the Department of Law, University of Turin, Italy. His main fields of interest are Artificial Intelligence and law, network theory, robotics, and information technology law. In his 30 years of research, he has published books, countless papers and participated in projects of the European Union. In a three-part interview series, we talk with him about what has been accomplished and what lies ahead. In part 3, he speaks about current issues in a global context.
Interview: Jonathan Mehlfeldt
Rechtverblüffend: Professor Pagallo, I would like to talk about the issue of affording legal personality to AI entities, as discussed in your book “Advanced Introduction to Law and Artificial Intelligence” that you co-authored with Dr. Woodrow Barfield. Could you give us an overview of the debate?
Ugo Pagallo: Well, I would first insist on the difference between the law and hard sciences, meaning engineering, computer sciences, mathematics and the like. Concerning the idea of agency, it is worth mentioning that there is no legal system in the world that has accepted the legal agency of intelligent machines. They are tools and they are regulated as such. On the other hand, it’s pretty obvious – at least nowadays, we discussed that a lot 15 years ago – that such machines are agents in the engineering-specific meaning of the word: They are adaptable; they are autonomous. Here lies a first source of tension between those producing the technology and those regulating the technology.
"There is no legal system in the world that has accepted the legal agency of intelligent machines."
On this basis, the idea of legal personhood for these smart machines has been debated by scholars for at least the past 30 years. The topic became increasingly relevant again after the famous – or infamous, depending on the viewpoint – document of the European Parliament suggesting that the European Commission examine whether it would make sense to set up new forms of “electronic personhood” for some of these machines. Since that report of Parliament, a huge debate has ensued. In my view, there was and still is a basic misunderstanding: When talking about legal personhood, we should distinguish between personhood and agenthood.
Rechtverblüffend: What exactly is the distinction?
Pagallo: My favorite examples are Roman slaves and EU diplomats. In both cases, you have examples of legal agenthood: As you know, slaves where the businessmen in ancient Rome, but they were not considered persons. If talking about slaves seems problematic, think about EU diplomats: For decades, the European Community had no legal personhood, yet the diplomats were, in legal terms, agents.
Notwithstanding the citizenship of Sophia the Robot [a humanoid intelligent robot that was awarded the Saudi-Arabian citizenship in 2017] – which, by the way, “sold”, so to speak, “her”, so to speak, self-portrait recently for more than 600.000 US-Dollars – it is my view that any idea of legal personhood, meaning constitutional rights, is an insult to most of the eight billion people living in this world who have no human rights. It is an insult to the people in Myanmar in this very moment.
Rechtverblüffend: Is there another approach that doesn’t involve legal personhood?
Pagallo: I do think that it makes sense to consider possible forms of legal agenthood, meaning that if something goes wrong, the robot might pay or cover the damages, rather than a human. But this does not mean – as many claim – that it’s a way to protect the “Fat Cats” in Silicon Valley in cases such as Alexa buying things on the Internet on its own – which already occurred, as you might know – and damaging the owner of this Alexa. It would be another insult to think that Amazon is off the hook here just because Alexa is an autonomous agent with its own legal personhood.
Rechtverblüffend: What do you have in mind when considering legal agenthood?
Pagallo: I am thinking of tackling the opposite scenario: Cases of a certain complexity of the social human-robot-interaction or cases of distributed responsibility, the latter being typical in environmental law. When many tiny, single decisions are put together to a final output, picking one single decision individually can be problematic. In both types of cases, it is very difficult to just pick someone and say: “Hey, you are responsible for the final outcome”.
We have a substantial problem when people suffering damages have no way of being protected. If there is no individual entity accountable for that final outcome, who will cover such damages? In those cases, it might make sense to think about new forms of artificial agenthood. But that’s the point: To me, this is an empirical issue; a tool to tackle a real problem. It is not, or at least not only, a philosophical exercise.
"If there is no individual entity accountable for that final outcome, who will cover such damages?"
Rechtverblüffend: Regarding the “Fat Cats” in Silicon Valley that you already mentioned, what role does law play in a globalized world with these powerful private actors emerging? How important is the Law of New Technologies and does it need to adapt?
Pagallo: “Fat Cats” have existed since the first industrial revolution – and we are currently at the fourth one. Rather than talking about Exxon or Shell, we are now just talking about Facebook and Google. So, on the one hand, nothing particular has changed. On the other hand, however, it can be very tricky to understand how to apply the old rules, envisioned for the previous “Fat Cats”, in the digital world. That explains why, for example, anti-trust-powers have remained “shy”, so to speak. They aren’t timid or shy themselves, but it can be tricky to apply the old idea of a market in the digital world. But especially in Germany or in the US, something is changing in this regard.
Rechtverblüffend: Another specificity has to do with the fact that platforms are instrumental for users.
Pagallo: When regulating ISPs – because, at the end of the day, we are talking about service providers on the Internet, although gigantic – we are also regulating billions of users. Facebook has 4 billion users; Apple has more than 1 billion. In many cases, there is a pathological dependency – at least in my view – of many of these users regarding such platforms. My old joke, especially to first-year students, is “Try not to use your WhatsApp for one hour”. The problem is that when regulating these companies, you are also regulating a relevant environment for billions of people on this planet. On top of that, we have to consider the issues regarding your previous question on legal regulation.
Rechtverblüffend: So there are many aspects as to why the law is so relevant nowadays.
Pagallo: Yes, and this is why we have had three really important acts here in Europe over the past few months: The Data Governance Act, the Digital Services Act and the Digital Markets Act. And even the Digital Services Act, which is more strict or “severe”, also follows a co-regulatory approach, rather than a simple top-down instruction.
Rechtverblüffend: Professor Pagallo, the last two questions: Does regulation impede innovation? And should it?
Pagallo: Yes, of course it does. There are many examples of regulations hindering technological innovation. A good example nowadays is how Italy has implemented the GDPR, especially in the health sector. Generally speaking, the GDPR leaves discretion to the Member States considering scientific research. The regulation of sensitive data processing in the health sector in Germany is different from the regulation in France, Italy and so on. In Italy, they have adopted a strict – if not too strict – regulation concerning whether and to what extent it is legitimate to process sensitive medical data for research purposes. This is an example of how strict regulations may hinder smart technological advancements.
And that brings us back to an issue I stressed before: The problem is to combine legal safeguards, the rule of law, our human rights, our fundamental rights and so on with the flexibility that the law must have in order to tackle technological innovation without hindering it and without requiring frequent revisions and amendments of the law while still keeping pace with technology.