[REPORT] Regulating Under Uncertainty: Governance Options for Generative AI
24 September 2024
In the wake of recent transformative events reshaping platform governance, Can Şimşek conducted an interview with Professor Kate Klonick, member of the Scientific Committee of the Chair Digital Governance and Sovereignty. The transcription of this fruitful interview below delves into how recent high-profile incidents—ranging from Elon Musk’s clash with Brazil to his support of presidential candidate Trump, his public insult directed at former EU Commissioner Thierry Breton, and the arrest of Telegram CEO Pavel Durov before the changes to Telegram’s privacy policy—interconnect and influence the evolving dynamics of the geopolitical and technological landscape. Buckle up for intriguing insights into freedom of expression, cyber-libertarianism, and platform governance.
We frequently hear platform CEOs invoke the concept of “free speech” when discussing or opposing content moderation laws like the DSA. Could this ongoing tension stem from the differing interpretations of freedom of expression in the EU versus the US? Or is this discourse around “free speech absolutism” purely political and sceptical in nature?
Kate KLONICK : Before I spent time in the EU, I really struggled to understand from a first-hand level, why there was so much seeming disagreement or why so many laws came out in the EU that we would balk at in the United States around freedom of expression. I really think it is more of a structural problem; it is a governmental and civil law structural problem and a normative problem. It’s best understood that way. What I mean by that is that the First Amendment as a constitutional matter bars the government from passing laws that could abridge citizen speech. The US courts have essentially interpreted that as the government cannot infringe almost any speech. We have hate speech, which is legal in the United States. Pornography, extreme pornography is legal in the United States. Burning the flag is legal in the United States. There are lots of things that are legal in the United States that might be illegal in Europe or elsewhere, obviously. What that means is that they’re protected citizen rights to speak back to the government. This is the main idea, and this comes from the idea that the thing you have to fear the most, and this is a very American thing. You have to really go back to what America was founded on. The government came after us. Like the crown and England came after the revolutionaries, and the crown is the enemy, the government is the enemy. But simultaneously, recognition of the founding fathers being that, of course, we need a government in order to operate. This is part of our civics education, this is part of how we understand ourselves as a nation, this is everything; it is a very cultural understanding. The government is what we’re afraid of the most. The government which has the guns which has the force which has the police power coming after us and restricting us. That is kind of behind both Americans being able to arm themselves with the second amendment and Americans being able to speak against the government.
On the other hand, there’s not a huge amount of difference in what you’re allowed to say in the EU. It’s not like you’re arrested for saying things in Europe. I’m constantly amazed that Europe as a whole is not more concerned about passing such open-ended regulation like the DSA that has the potential enforcement power to really restrict what companies are able to do but also what citizens are able to do on platforms. Two generational memory of World War II and an autocratic government coming to power would seem to me like the exact type of situation that the Americans are fearful of, which is that you have a government that spies on its citizens, you have governments that become autocratic in nature and then use the laws against their citizens. I think that kind of sets the bar at a very basic place of why there’s such disagreement. But I agree with you. At the end of the day, there’s not so much difference.
I believe highlighting the historical roots of freedom of expression is indeed crucial. However, technological progress is adding an entirely new dimension to this issue. The scale is immense, especially with global platforms like X. For instance, incidents involving insults directed at an EU Commissioner or controversies in Brazil are amplified, with far-reaching effects on a worldwide scale. Given this context, how should we approach and understand these issues in today’s world?
Technology changes the basic fundamentals of the people versus the government. Technology added a new node. This is an idea from Jack Balkin at Yale Law School. He has a paper called “Free Speech is a Triangle” which I think is a wonderful heuristic for understanding this. Especially for the United States. It was a dyadic relationship between citizens and government around speech; the government had the boot of the State, and people could push back with vote and voice. Platforms add a third node right? They make it a triangle. This is why you have tech optimism and this is also why you have tech pessimism. All of a sudden, you have a new node. You can go to this third party who can look at the citizens for you, and it also means that you can run around censorship because you have this third party that can allow you to get around the government. Essentially, this is a very natural idea for the United States in particular, with its First Amendment structure. What the government can restrict leaves a ton of space for private entities, be they press platforms, be they private companies, be they civil society groups, academics, things like that, leaves a lot of room for spaces to be created for speech and for places to be created for speech that are privately run, that are not the government. And so, like if you own a restaurant, you can decide what the speech rules are in your restaurant because you’re not the government. You can ban hate speech in your restaurant, you can ban hate speech on your digital platform if you want to. You have that right. This is why there is a lot of discussion about how the platforms have their own first amendment right as “private” platforms. I think this is a very natural idea because they have a platform in a way that they want. It is the right to create this platform that’s protected against the government. That is what my work has been so far about, is understanding what is happening in that third node and trying to understand what’s happening in that private governance space and organising all of that into concepts. What are the trends, how is this happening? Understanding it anthropologically, understanding it as a semi-legal structure. It’s almost like a type of legal structure that has taken place.
Both Durov and Musk occupy that third node and what their concerns are in very different ways are control and response to the government and their so-called concern and response to the citizen users on their site. It is a different kind of game, where both the politics of the situation and the economics of the situation is at stake. The legal technocratic option, like the European Commission coming in, is just another manifestation of a way of life, a little flex of power in order to be able to continue this kind of constant struggle between these three parties.
Are we witnessing a shift in the way these third nodes are acting? I believe we’re seeing a fragmentation of social platforms based on the political views of their owners. Before Musk, Twitter seemed like a public space which was rather neutral. Has Musk altered this dynamic or were platforms always curated by their CEOs?
It is super complicated what we mean by public and private because when we say public we mean that in a widely published and disseminated sense. In the U.S. we use the term public to mean government run. Then also having a “publicly traded company” means that you’re on like the New York stock exchange and beholden to the laws of the government and you’re beholden to a board and you’re beholden to your shareholders. There’s a whole other set of obligations and now, really complicatedly, for governance purposes and legal technocratic purposes, X is privately owned and no longer a publicly traded company; it is a private platform that publicly disseminates information to individuals. So, yes, from a communications or media standpoint, I think your impression of X as a public space is like a public platform.
What most people would say and what a lot of my work has shown is that we were actually kind of following Mark Zuckerberg’s whims, and we were at the whim of Jack Dorsey, before Elon Musk, and we were at the whim of like whoever was in charge of Alphabet or Microsoft or whatever else, or Cloudflare Matthew Prince. The reality is that we were; they just weren’t very transparent about it. Now, Musk has done a kind of a figure that you wouldn’t believe in. History has come out of this kind of well… I feel like there has been this decline in the stock of Elon Musk and I don’t mean that literally like all of his companies. He was seen as this absolute genius and now he’s just someone who posts memes on twitter and bad jokes to his followers. I think this is such an interesting moment because to your point, I and others were saying for years that the reason there needed to be internet governance is; listen, what you don’t realise is this is all at the whims of a few people in Silicon Valley.It is just a few people, and there’s no process for it, that’s to the people; and there’s no system of governance for these quasi-public spaces, these “public spaces” as you say feels public; there’s no type of responsiveness, there’s no democratic structure for people or individuals to speak back to the people who are making the rules that govern these private spaces. At a certain point it tips over to being so incredibly public that it becomes really an essential part of your freedom of expression. There needs to be a response to the people and there needs to be responsiveness in a system. CEO’s can become the law of the platform and that is exactly the problem.
It is a classic psychology bias to think that because you’re seeing something it just started happening. This is actually not true. Before this was happening for years and years and since the beginning, but Elon has shown us that it really is happening.
It is, in a way, good that Musk is not being discreet about this. People are becoming more aware of this situation, right? And this responsiveness issue brings us back to the Pavel Durov case. This situation isn’t primarily about encryption—since Telegram is not fully encrypted—or about moderation policies. Instead, it revolves around a preliminary criminal investigation. How do you interpret Durov’s arrest and the subsequent events?
Absolutely, so I think both Elon and Durov at a fundamental level, are two larger than life megalomaniac human beings who kind of think that the tech platform as a whole is their opportunity to voice or actualize, kind of their libertarian notions of thumbing their nose at government; that’s not a new thing. We experienced an internet that was born of a techno-utopian moment, like the John Perry Barlow idea of ‘we’re going to screw the government and let the government be screwed’; this is the moment for us to seize our rights as individuals. The internet is a hellscape if you do not have some type of rules and regulations. The internet would be illegible without governance. You need to organise public conversations, you need to have people see what they want to hear, they don’t want a fire hose of information. Even creating an app is a structure. It is an infrastructure that you’re creating to invite people to have some type of legible conversation.
Let’s just talk about child sexual abuse material (CSAM) for a second. CSAM is the worst of the worst of the internet. It has been one of the first things that came onto the internet. People will try to find a way to route around the government which bans very strongly this type of material and it’s criminalised speech so it is not protected by the first amendment. The development of PhotoDNA was a huge step forward in this. We are able to identify this very quickly through platforms. Nobody that’s not themselves a child sexual abuser is a proponent of CSAM. It is a very easy thing for everyone to agree on at a normative and cultural level. It’s not worth going to the mat for child sexual abusers, right? In a fight, no one wants the child sex abusers to win. We’ve just decided that it’s criminal activity.
This became very very easy to comply with child sexual abuse material laws because of technological advances like PhotoDNA. So what you have is most places on the internet complying with child sexual abuse material laws.
Telegram, as you point out, is not an entirely encrypted platform. It’s a multi-faceted platform with multi-faceted ways of communicating very loudly and publicly, and very privately, and in different types of ways. You have like a messages type function, secret chats which are supposedly encrypted though they’ve never posted their key and told you how they do it, that is the part that is the most encrypted and they’ve really been protective of. But there’s public broadcasting elements the way there is on Facebook Live, and there’s other types of group chats that are non-encrypted. Durov has made his entire political enterprise, his entire selling of the platform as a space of freedom. This libertarian thing is just marketing. Durov has made his stance because he sees the bigger threat being governments using platforms to surveil citizens. That’s what his stance has been: we’re not going to comply with subpoenas; we’re not even going to comply with CSAM requests or have any reporting things because we see that as the slippery slope to complying with any type of request. Now, to be fair, a lot of platforms are dealing with that slippery slope; it’s very complicated and it requires a lot of tech and hiring out. This creates a game of chicken with governments around the world that really want to prosecute him for running a platform that keeps up CSAM, the worst of the worst, and terrorist extremist content. Durov flying into France and landing there and kind of surrendering and kind of saying “okay I’m going to push this issue a little bit.” It’s literally a game of chicken. Then we’ve seen telegram revise their entire stance on things in the last couple of weeks. So this is at the end of the day a very interesting moment. As I said before, the three nodes of power are in play. The fact that the government could force one of those nodes to capitulate, even though I agree that Telegram is a terrible app and it does a lot of terrible things, can be seen as a strike against user rights and user options and user freedom.
When considering the power of platforms, it’s worth noting that they often possess additional leverage. For instance, platforms can potentially threaten to withdraw from markets or challenge regulatory bodies like the EU if their demands aren’t met. Do you think the libertarian rhetoric used in marketing by these platforms holds any validity? Do they fight back or do they seek to compromise behind the scenes?
More efficient bargaining is the backroom bargaining, it’s preventing all this from happening. You have this in almost every government in the world. There’s lobbying from different special interest groups and different individuals, and you get things changed before they become law or you change what they mean after they become law, or you have a settlement that happens. All of this is expensive. I keep calling it a game of chicken. The game of chicken is to keep calling bluffs, it’s to keep being the one who’s finally going to get hit by the car, and it’s unclear how that will play out. I mean at some point, economics is the biggest driving factor in this, which is why, like you can say all you want, but like economics or spending years in prison when you’re the CEO of a company are going to be the levers that are going to be the strongest sticks. We’ve seen a lot of reporting on this in the last couple weeks, the number of fines that are coming out of the EU around not just obviously the DSA but like all kinds of court decisions, the antitrust suits etc. It’s not that they’ll stand on principle and pull out of Europe. But maybe it won’t be profitable to operate in Europe. Digital governance in general is so interesting. It is just like a three-dimensional chess game. You’re constantly going between the threat versus reality of enforcement, and then, you’re constantly going between the power of law and the very real economic conditions that are the driving factors of for-profit companies, and then kind of these pie-in-the-sky ideals of megalomaniac CEOs and what they’re willing to risk and what their lawyers are telling them they’re willing to risk.
What’s so interesting to me as a person who studies this is the timing of the Durov and the Musk moments. The timing of the Durov France moment in which the civil libertarians were like “can you believe that they’re coming after a CEO! This is bad for free speech” and found themselves semi-championing Durov, but then for like the worst type of speech imaginable. I think that most people can say that Telegram should comply with things like CSAM and the rule of law in the countries in which it operates, but that becomes a really hard question once you’re going through it and you’re in autocratic countries and places that don’t protect their citizens. You could see Durov as either like a villain or a hero and then the same is true of Elon Musk. That’s the question in Brazil. This is a soft democracy, and has a lot of different types of norms. We have the exact same type of situation here, which is that you’re having a few judges in both France and in Brazil, having independently issued these orders against these platforms and against these individuals. And you have Elon Musk standing up to Brazil. Is he a hero of free speech or should he be just following the law? What we’ve spent very little time talking about is what individual consumers want and whether or not they approve of all of these government requests for subpoenas or what the platform’s doing or not doing? There’s no collective action mechanism to capture that. The tail that’s wagging the dog is economics and politics, and it is not at all the voice of individual consumers and citizens which rely on these platforms.
Kate Klonick is an Associate Professor of Law at St. John’s University Law School (New York), Affiliate Fellow at the Information Society Project at Yale Law School (Connecticut), Nonresident Fellow at the Brookings Institution; during the 2022-2023 academic, she wasVisiting Scholar at Harvard University’s Rebooting Social Media Initiative (Massachusetts).