The New Social Contracts

Contracts rule the digital world. Today’s terms of service dictate Americans’ speech rights, privacy rights, and even Fourth Amendment rights. This makes terms of service something like public law for the internet. In a way, this is not terribly surprising. From the beginning, the American approach to internet regulation was designed around freedom of contract. As President Clinton put it, way back in 1997, U.S. internet policy would be built around “a decentralized, contractual model of law,” unlike Europe and China with their top-down rules. 

A quarter of a century later, one would expect a huge body of scholarship and case law interrogating the huge social impact of online contracting. But there is shockingly little. To be sure, there is a great deal of scholarship about the new digital contracting landscape. But the central focus of that research has been on the procedural defects of clickwrap—most importantly the idea that when you click “I agree,” you are not in fact reading the contract, and so there is no genuine mutual assent. Clickwrap is an interesting puzzle, but it is a side issue. Rather, the notable thing about terms of service is that they are effectively the ground rules for our digital society, supplementing and even supplanting our constitutional norms. And as best I can tell, this development has been largely ignored by contracts scholars. 

How can this be? This article tries to unpack the problem. (It is drawn from a longer article.) First, I outline the ways in which contract is a powerful force today—effectively acting as the public law for our internet. Then I show that contract law and scholarship have had surprisingly little to say about this. Finally, I argue that we face a choice between two imperfect alternatives: We can retreat from our contractarian internet policy and impose limits on contract ex ante, or we can ask the common law of contracts to step up and police the public harms of our new social contracts. Neither option is great.

Contract Law as Public Law for the Internet

Today’s platform contracts determine all manner of user rights, including speech rights, privacy rights, and more. This is partly a function of U.S. public laws, which often defer to the private market. And it is partly a function of our contract law, which suggests that users can contract for just about anything, including opting into a system of private governance like the digital fiefdoms that rule today’s internet. 

Who determines free speech rights online? A wide array of commentators such as Kate Klonick, Evelyn Douek, and Daphne Keller have shown that the speech rights of users turn in large part on the content moderation policies of the platforms, which users agree to as a condition of the service. It is not surprising that the purveyors of a digital service would create rules about what kind of content is allowed there. What is surprising, however, is that as a society we have migrated most of our communications to these platforms, and we largely prohibit the state from interfering in their content moderation. The stunning result is that a few platforms set the on-the-ground speech rules for our society. 

This is of course in part a result of the First Amendment, which appears to be hostile to just about any conceivable government intervention in the digital market. The First Amendment is unusual, however, in its sweep and recent embrace of private enterprises. Perhaps other constitutional norms are not so deferential to private contracts. What about the Fourth Amendment, which sets the limits on when the state can search or seize your private information?

Orin Kerr recently showed that even though there is nothing in the Fourth Amendment to suggest it, courts widely view Fourth Amendment rights as something that can be waived by contract. In a range of cases involving police access to cloud storage, courts held that users had waived their reasonable expectation of privacy—the linchpin of their Fourth Amendment rights—by agreeing to the terms of service. In these cases, platform contracts seem to override constitutional rights given by the state. 

In a range of issues from speech to discrimination to privacy, the ground rules for our society are set out in private contracts between market actors. None of this would matter if the platforms represented a small segment of society, but they have become, as the Supreme Court put it recently, our “modern town square.” This makes terms of service weird as contracts. First of all, unlike almost every other commercial contract, these contracts are society-wide. Between WhatsApp, Facebook, and Instagram, Meta is in an ongoing contractual relationship with over 80 percent of the country. One is hard-pressed to think of a commercial contractual relationship like that in the pre-digital era. Second and more important, unlike commercial contracts these terms of service set out the rules for society. To be sure, they are commercial contracts that outline the rights and responsibilities of parties to a market exchange. But they also outline how and where and why the platform can sensor political speech, they explain how and why the platform can share your private information with the police, and they set the basic terms of our digital world. They are community rules.

Taken together, today’s terms of service have the sweeping ability to provide society with the kinds of rules that are typically set out in our public laws. 

Contract Law Says Nothing

Surprisingly, the common law of contracts has very little to say about this. 

Part of this has to do with the common law itself. Contract claims are typically brought by a single party to a contract. Yet the social impact of today’s digital terms of service is felt in the aggregate. It is hard to imagine an individual claimant having the incentive or standing to bring a suit that would require a court to inquire into the widespread social impact of an agreement. Part of this has to do with contract law’s general focus on procedural questions, rather than substantive questions. For the most part, contract law is interested in whether the parties entered into the deal freely, not whether the deal itself is a good or fair one. And part of this has to do with contract law’s general lack of interest in third-party harms. As scholars like Aditi Bagchi, Dave Hoffman, and Cathy Hwang have argued, contract law is much more attentive to the parties’ interests than those of third parties, including the public. 

To be sure, scholars have been deeply worried about the rise of boilerplate agreements including clickwrap and the general sense that no one is reading these agreements. Take a look at any contracts casebook: Terms of service are covered under mutual assent. The core question is whether something like a “meeting of the minds” can occur in a world in which terms of service are a nuisance and most of us just click “I agree” as quickly as possible. 

But the problem is not mutual assent. The technology firms have been prodded to make their terms of service clearer and simpler, and they have done so. Go read the terms of service for a major service like Instagram; they are easy to understand, written in accessible and clear language that is not meant to obscure. The problem is not that Americans do not read contracts. The problem is that they are fine with the terms. The problem is that the U.S. has decided, as a society, to opt into a private empire, where public laws do not apply.

That is a problem that contract law does not have much to say about. 

Moving Forward

Essentially, I think the U.S. is in a bind. Americans want the internet to comport with the basic rules of society, consistent with our laws and broader public values. They do not simply want the internet to be an unregulated free market. If the U.S. wants rules, and they aren’t to come exclusively from Mark Zuckerberg and Elon Musk, we need to do one of two things: We can retract from the contractarian model of internet governance; or we can allow contracts to set the rules, but courts need to get in the action and police the societal impacts of these agreements.

Neither of these is a great option. That’s the bind. The first option means ex ante, top-down rules where the state legislates limits about what can be contracted. Imagine a privacy statute that limits who can contract away what private information. Ex ante sounds nice, but who is to say what is the right rule this early in our tech transition? And the legislature is a mess. It’s fractious; it’s captured. And for all the noise that is made every day about this or that issue—witness the huge public hearings recently on protecting children online—Congress has not done much of anything for decades. As a result, like many of my colleagues in the law and tech space, I am not enthusiastic about the prospect of sweeping legislation.

The second option means ex post judicial intervention in platform contracts. This is my ugly, second-best solution. I teach contract law, and every year I walk my class through the horrors of having an unelected judge second-guess the judgment of two private individuals. It is strange for me—for anyone who loves contracts and contract law—to argue that courts ought to pursue a line of inquiry that has nothing to do with party intent. But as I noted above, these are not normal contracts. They are society-wide pacts, and they have enormous social impact. The state ought to be in the business of at least making sure they are consistent with our most basic societal values. 

And there is a doctrine for doing just that. The public policy doctrine—much dismissed by scholars, famously unruly—asks courts to evaluate whether a private deal is consistent with the laws and policies of the jurisdiction. So even if two private actors, or three, or 300 million, all wanted to sign on to a private agreement that undermined the First Amendment, or the Fourth, a court could invalidate it. That, I think, is a promising avenue for some limited judicial review of today’s social contracts. 

There is still the matter of standing. None of Facebook’s many users—or TikTok’s or Google’s—have the incentive or standing to bring a suit on behalf of the nation. If the problem is aggregate societal harms from many individual contracts, then the solution cannot be individual complaints from single litigants. The solution must entail some kind of aggregated complaint that alleges harms on behalf of the entire republic. And it must allege some kind of public harm. Here is where I think state attorneys general could play a role. They could bring suit in parens patriae—on behalf of the sovereign’s citizens—to block the enforcement of today’s platform terms of service wherever they are inconsistent with public policy. 

As I say, this solution leaves me very uneasy. It does not solve all or even most of the regulatory problems posed by our new digital world, where we live much of our lives in corporate walled gardens. But it is a start. It is a simple backstop that ensures that the state is occasionally asked to confirm that our massive online contracting is consistent with the basic public values of society.