What Carpenter Tells Us About When a Fourth Amendment Search of Metadata Begins.

• Author: Ramirez, Geneva

"Given the advancing state of both the remote sensing art and the capacity of computers to handle an uninterrupted and synoptic data flow, there seem to be no physical barriers left to shield us from intrusion." (1)

CONTENTS INTRODUCTION I. METADATA A. Types of Metadata and Its Uses B. CSLI C. The Aggregation Problem II. WHAT CARPENTER TOLD US A. The Facts B. The Supreme Court's Reasoning 1. Expectation of Privacy in Physical Location and Movements 2. Expectation of Privacy in Information Shared with Third Parties 3. What Carpenter Tells Us About When the Search Begins III. STAGES OF DIGITAL SURVEILLANCE USING METADATA AND A RESULT-ORIENTED APPLICATION OF FOURTH AMENDMENT METADATA ANALYSIS IV. WHEN THE SEARCH OF METADATA BEGINS: A FRAMEWORK FOR ANALYSIS A. Identifying the Information Sought B. Determining Whether That Information is Subject to a Reasonable Expectation of Privacy C. An Exception to the Result-Oriented Model and How the Third-Party Doctrine Applies V. WHAT THE WARRANT MUST CONTAIN CONCLUSION INTRODUCTION

The Fourth Amendment protects individuals from unreasonable government searches by requiring the government to procure a warrant supported by probable cause that "particularly describe[es] the place to be searched, and the persons or things to be seized," prior to searching persons or their homes, papers, or effects. (2) Traditionally, the Supreme Court's conception of the Fourth Amendment's protection was tied to physical space. The government could not seek information in an individual's constitutionally protected areas (e.g., one's home or office) without first acquiring that individual's permission or obtaining a warrant. (3) This made sense because people stored private information in private spaces, and private and public spaces were usually delineated by clear, physical boundaries. But as technology has advanced, the constitutionally-protected-area conception of the Fourth Amendment fails to provide adequate protection for that private information that is now stored, communicated, and generated digitally.

Recognizing this technological shift in 1967, the Supreme Court in Katz v. United States (4) adopted the view that "the Fourth Amendment protects people, not places." (5) In doing so, the Court laid down a new reasonable-expectation-of-privacy test that is not confined to constitutionally protected areas or tangible things, but extends outside of the home and office to cover electronic and digital information and other communications to which an individual has a reasonable expectation of privacy. (6)

Today, over fifty years after Katz, the reasonable-expectation-of-privacy test has become an even more vital safeguard against unreasonable government intrusions that are due, in large part, to the Internet. As one observer noted, "the Internet is not some standalone, separate domain where a few of life's functions are carried out .... Rather, it is the ... place where virtually everything is done. It is ... where the most private data is created and stored." (7) In fact, "more data has been created [in the years 2014 and 2015] than in the entire previous history of the human race." (8) And the amount of data that is created and stored on the Internet continues to grow exponentially: by 2020, "about 1.7 megabytes of new information will be created every second for every human being on the planet." (9)

The data we produce using digital devices fall into two categories: content--the substance of digital communications and activities--and metadata--data about content. (10) While the content of our digital activities is generally agreed to be subject to Fourth Amendment protection, (11) some have argued that metadata is not. In 2013, following Edward Snowden's revelation that the National Security Agency was collecting metadata about U.S. citizens on a massive scale, Senator Dianne Feinstein argued that metadata is not protected under the Fourth Amendment. (12) She reasoned that, because the records being collected did not include content, names, or locations, their collection did not qualify as surveillance. (13)

But while it is true that metadata on its face is almost meaningless to most people, it can, when analyzed, reveal the most intimate details of our lives. Our digital-activity records create information-rich "metadata trails" that, because of metadata's structured nature, can "often yield information more easily than ... the actual content of our communications." (14) And the greater the quantity of metadata analyzed, the more revealing it can be. Large datasets can be used for everything from mapping an individual's location over a period of years, to identifying a person's relationships, habits, and behaviors. (15) This previously inaccessible source of information has the potential to make the government practically omniscient--and it is widely agreed that "police omniscience is one of the most effective tools of tyranny." (16)

But to hold the government accountable for analyzing our metadata, we must revisit one of the Fourth Amendment's fundamental questions: When does the search begin? (17) This question is important because it determines not only at what point in the process of metadata analysis---i.e., acquisition, analysis, or use (18)--must the government secure a warrant, but also the ways in which the government may analyze and use our metadata after acquiring it. (19) While the Supreme Court has not directly addressed this issue, it provided some guidance in Carpenter v. United States. (20)

In Carpenter, the FBI subpoenaed 127 days of Timothy Carpenter's cell-site location information ("CSLI")--metadata in the form of time-stamped records containing the identification number of the cell site that his phone was connected to at a given time--from his network provider. (21) The Court held that "[t]he location information obtained from Carpenter's wireless carriers was the product of a search"; (22) thus, the government had violated the Fourth Amendment by failing to acquire a warrant. The Court based its holding on its reasoning in two lines of cases. (23) The first concerns a person's expectation of privacy in their long-term physical movements. (24) The second concerns "a person's expectation of privacy in information voluntarily turned over to third parties." (25) Though the Court did not directly answer the question of when a search of metadata begins, its reasoning and the concerns it considered provide a foundation for determining when such a search begins. The answer to this question determines what the Fourth Amendment protects--metadata itself, or the information that results from its analysis--as well as how the government may use metadata it has legally obtained.

This Note explores when a Fourth Amendment search of one's historical metadata begins, based on the Supreme Court's reasoning and concerns in Carpenter. Part I provides an overview of what metadata is and what types of information its analysis can yield. It then takes a closer look at CSLI--the metadata at issue in Carpenter--and addresses the threat to privacy posed by the aggregation of large amounts of metadata. Part II provides a close analysis of the Supreme Court's reasoning in Carpenter. Part III discusses the various stages in the process of metadata-based surveillance at which a Fourth Amendment search could begin. It argues that, to best safeguard privacy, the search must begin not, when metadata is acquired, but when it is analyzed to reveal private information. Part IV proposes a framework for analyzing when a metadata search begins. Finally, Part V suggests a reinterpretation of the Fourth Amendment's particularity requirement as it pertains to historical metadata.

1. METADATA

   This section explains what metadata is, the uses for which it is collected, and what it can reveal when analyzed. It then provides an in-depth explanation of CSLI -the metadata at issue in Carpenter--and continues by describing the enormous range of private information that can be gleaned from the analysis of aggregated metadata.

   1. Types of Metadata and Its Uses

      Metadata is data about data. It does not describe a communication's or digital activity's content, but instead it comprises information about that communication or activity. For instance, if you call your mother, the metadata about the call will not take the form of a transcript of the conversation. Instead, it will contain the call's length, the date and time when the call took place, both the initiating and the receiving telephone number, the cell-site identification number of the cell tower your phone was connected to, and other logistical data. (26) Basically, it. is just a list of numbers. Almost every digital activity we engage in leaves behind "rich metadata trails." (27) And virtually everything we do while browsing the Internet or using digital devices is recorded and stored, "creating a permanent record of unparalleled pervasiveness and depth." (28)

      This collection of data about our communications and activities may seem innocuous compared to the collection of records detailing those communications and activities. But when our metadata trails are analyzed, metadata can reveal an intimate picture of our lives. (29) In fact, by virtue of metadata's structured nature, metadata analysis can often reveal those details more easily and cost-effectively than the content of our communications. (30) This structure makes it easy to store and quickly analyze vast sets of data for patterns that can reveal our "personal details, habits, and behaviors." (31) By contrast, to analyze the content of a phone call, an analyst must transcribe the conversation, determine its meaning (taking into account a multitude of factors including language differences and code phrases), and identify relevant information. (32) The government simply does not have the resources to perform content analysis on the phone calls of three-hundred million Americans. (33)

      Although...