Friday, August 25, 2023
Orin Kerr on Terms of Service and Fourth Amendment Rights (Once More, with Feeling)
I flagged Orin Kerr's Terms of Service and Forth Amendment Rights as something I intended to read and write about back in February. I wasn't sure when I was going to get back to it, and today is the day. The article is now forthcoming in the University of Pennsylvania Law Review (congratulations all around!), but if you can't wait until then, you can find the draft on SSRN.
The article explores the possibility that Terms of Service (ToS) might allow large corporations to turn your personal data over to the government, thus providing the government with a ready alternative to the Fourth Amendment's warrant requirement. As Professor Kerr puts it, "What the Supreme Court has given, Terms of Service might take away."
Courts are divided. Some will not allow people to contract away their rights by agreeing to ToS; most treat use of the service as a waiver of Fourth Amendment rights. Professor Kerr argues that the courts that treat ToS as a waiver get things wrong. He looks at four doctrines where one might think that ToS might lead to lost privacy and concludes that none of them should affect Fourth Amendment rights. An essential step towards this conclusion is Professor Kerr's insistence that ToS clarify relationships relevant to Fourth Amendment doctrines, but those relationships, not particular terms, govern Fourth Amendment rights.
Professor Kerr analogizes breach of ToS to breach of a car- or apartment- rental agreement. A user's breach of those agreements has no impact on Fourth-Amendment rights. However, he acknowledges, ToS invite a difference line of reasoning: Fourth Amendment protections require rights; ToS define rights in online accounts; therefore, ToS define Fourth Amendment rights in online accounts. Actually there are two possible routes to the same conclusion, Professor Kerr notes. In addition to treating contracts rights as a waiver, in cases in which a private company voluntarily turns over incriminating evidence to the governments, courts sometimes avail themselves of the third-party doctrine. The search was not conducted by the government, and the Fourth Amendment does not protect against private searches.
Professor Kerr thinks that courts that think ToS largely irrelevant to the Fourth Amendment analysis get things right. The law has never recognized a diminishment of the reasonable expectation of privacy just because phone companies have access to user data, or landlords or cleaners have access to an apartment. The tough cases arise where the ToS have language clearly indicating that the provider will share incriminating information with the government. To some courts, such language renders the ToS distinguishable from a typical lease of a car, a hotel room, or an apartment.
Professor Kerr then makes four compelling arguments for why existing caselaw forecloses arguments that would permit the government to use ToS to get around the warrant requirement. First, the caselaw on shared spaces provides that agreeing to share an office or an apartment with others does not deprive one of a reasonable expectation of privacy vis a vis the government. Second, formalizing rules of access to private spaces, as one might do in a hotel, business, or apartment rental agreement does nothing to change one's expectation of privacy against the government. Third, this is so, even if the user violates a rental agreement. So, for example, police cannot carry out an illegal search of rental car driven by an unauthorized driver even if that amounts to a breach of the car rental agreement. The same is true in cases involving breaches of apartment leases and use of hotel rooms. Finally, the rule for shared space operates differently in government spaces, but Professor Kerr notes, that exception to the expectation of privacy has always been limited to government offices.
Professor Kerr then makes a remarkable discovery: the cases that treat agreement to ToS as carving out an exception to traditional Fourth Amendment doctrine rely on the government spaces cases! They mistake the exception for the general rule.
Finally, Professor Kerr discusses four doctrines where one might think that ToS could make a difference. They are the private search doctrine, third-party consent, direct consent, and abandonment. Professor Kerr thinks courts should treat ToS as irrelevant to all four doctrines.
Here, I must admit, the argument loses me. Either I don't understand it, or it gives me little comfort. With respect to private search doctrine, the case on which Professor Kerr relies, United States v. Rozenow, seems to allow Internet Service Providers (ISPs) to share information with the government when their motivation for investigating an account is violation of their ToS. Rozenow was treated as having no expectation of privacy in the material on his account because it was discovered through a private search motivated by his ISP's independent interest. It sure seems like the ToS very much matter in determining what actions an ISP might take, beyond suspending an account, including storing incriminating material or sharing it with law enforcement.
In discussing third-party consent, Professor Kerr acknowledges that ToS might be written so as to track the doctrine and allow private companies to share incriminating data or information with law enforcement. However, he argues that the inquiry hinges not on theToS but on the reality of shared access to a user's data. If the ISP doesn't actually access the information, the expectation of privacy survives. I would argue the opposite. Companies put language in their ToS about shared access to data because they use our data. That's whey they don't charge us; we pay by sharing our data with ISPs. Our data could not be shared with law enforcement if there were no ToS evidencing the users' consent for ISP access of and use of the data. It seems to follow that in many contexts, the third-party consent doctrine might provide a way around the Fourth Amendment's warrant requirement.
In discussing direct consent, Professor Kerr argues first that consent to ToS cannot be construed as consent to a government search. The connection is too attenuated. I am happy to defer to Professor Kerr's vastly superior knowledge of Fourth Amendment law here, and I find it entirely believable that a more stringent definition of consent should apply in the Fourth Amendment context than in the contractual context. His second argument is less convincing: ToS cannot bind users because nobody reads them. Isn't it pretty to think so?
Finally, his argument on abandonment references his earlier arguments. He makes the strong point that it is hard to construe registration of an account as an abandonment of the expectation of privacy in that account. That might be convincing but for the ToS that might provide that by using an account you are agreeing to share your data with the ISP and abandoning any expectation of privacy in that data.
Despite these critical comments, I hope that Professor Kerr's way of thinking about these things succeeds. Nancy Kim and I expressed our concerns about Internet Giants as quasi-governmental actors nearly a decade ago. We raised some of the same concerns about consent that Professor Kerr does. But we adopted a pessimistic tone. Professor Kerr's work could provide more grounds for optimism.
More generally, I have been posting in this space periodically about interactions between contracts and the First Amendment. My view of the caselaw is that courts in certain contexts have allowed people to contract away Fourth Amendment protections, but they sometimes treat First Amendment protections as ironclad, even when the expression at issue is not core political speech, notwithstanding contractual interests. Professor Kerr's approach would bring the jurisprudence on the relationship of contracts and the Fourth Amendment more in line with that on the relationship of contracts and the First Amendment.
https://lawprofessors.typepad.com/contractsprof_blog/2023/08/orin-kerr-on-terms-of-service-and-fourth-amendment-rights-once-more-with-feeling.html