Sixth Circuit creates circuit split on private search doctrine for computers

The Sixth Circuit handed down a new decision on computer search and seizure that may be the next computer search issue to make it to the Supreme Court. The issue: How does the private search reconstruction doctrine apply to computers? The new decision creates an apparent circuit split with the Fifth and Seventh Circuits.

First, some context. Private parties acting on their own are not regulated by the Fourth Amendment. When a private party has conducted a prior search, the private party can reconstruct the search for the government without implicating the Fourth Amendment. The idea is that the private search has already shed the Fourth Amendment protection over what was searched, so that the government can ask the private party to redo the private party's search to show the police what the private party saw. Granted, if the officer asks the private party to conduct a new search that goes beyond the old one, then that can violate the Fourth Amendment. The new search that exceeds the old one is at the government's behest, and the new invasion of privacy triggers the Fourth Amendment. But just going over the old ground is permitted.

The idea seems simple enough. But the application raises a puzzle: When a private party sees a file on a computer, what exactly has been searched for purposes of later reconstruction? I discussed this problem in my 2005 article, Searches and Seizures in a Digital World. The question is, what's the right measuring unit to use—the data, the file, the folder, the physical device, or something else?

The issue is really important for computer searches, as it determines how much the government can search computers without a warrant after a private citizen finds evidence of crime on a computer and calls for help. The cases were already mixed in 2005, although at the time the Fifth Circuit was the only federal circuit court to weigh in. The Fifth Circuit had held that the unit was the physical computer, so that a private search of one file allowed the private party to turn over the entire computer to the government for a warrantless search.

Since then, there have been some added cases. In 2012, the Seventh Circuit joined the Fifth Circuit by adopting the unit of the device. And last month, a cert petition was filed at the Supreme Court on this issue in Gunter v. United States. But I hadn't thought there was a particularly clear split. At least until this morning.

This morning, the Sixth Circuit handed down a new case, United States v. Lichtenberger, that adopts the proper unit as data or a file instead of the physical device.

The facts: Girlfriend Holmes suspected that boyfriend Lichtenberger has child pornography on his computer, and on her own she searched his computer and found a bunch of child pornography images. Holmes contacted the police. An officer came to the home and asked Holmes to boot up the computer and show him the child porn. She booted up the computer, entered in the boyfriend's password, and "opened several folders and began clicking on random thumbnail images to show him." When the officer saw that some of the images were child pornography, he told her shut down the computer. A warrant was later obtained to search the computers based on what the officer had seen. But here's the critical fact: Holmes later testified that she wasn't sure that the files that she showed the officer were the same ones she had seen. In other words, it was clearly the same physical device that had been searched earlier. But it wasn't clearly the same files.

The Sixth Circuit suppressed the images, ruling that the Fourth Amendment was violated when the girlfriend opened up new files that she hadn't seen before at the officer's direction. The opinion was by Judge Donald, joined by Judges Merritt and Stranch.

First, the Court considered whether the private search doctrine could apply at all because the search applied in the home. There's actually a circuit split on whether the private search doctrine applies in the home—indeed, here's a New Jersey Supreme Court case jumping into the split handed down just yesterday. The Sixth Circuit has held that the private search doctrine doesn't apply to search of a home. But today's decision rules that this doesn't apply to a search of a computer, even if a computer is in the home. So the private search reconstruction doctrine applies.

The question is, how? Here's the court's analysis:

We find that the scope of Officer Huston's search of Lichtenberger's laptop exceeded that of Holmes' private search conducted earlier that day. This is, in large part, due to the extensive privacy interests at stake in a modern electronic device like a laptop and the particulars of how Officer Huston conducted his search when he arrived at the residence.

We evaluate "[t]he reasonableness of an official invasion of the citizen's privacy . . . on the basis of the facts as they existed at the time that invasion occurred." Jacobsen, 466 U.S. at 115. Under the private search doctrine, the critical measures of whether a governmental search exceeds the scope of the private search that preceded it are how much information the government stands to gain when it re-examines the evidence and, relatedly, how certain it is regarding what it will find. Id. at 119-20 (finding the DEA agent's search permissible because "there was a virtual certainty that nothing else of significance was in the package[.] . . . The advantage the Government gained thereby was merely avoiding the risk of a flaw in the employees' recollection, rather than in further infringing respondents' privacy."); see also id. at 117 ("This standard follows from the analysis applicable when private parties reveal other kinds of private information to the authorities. . . . The Fourth Amendment is implicated only if the authorities use information with respect to which the expectation of privacy has not already been frustrated.")

The Court then argued that "[s]earches of physical spaces and the items they contain differ in significant ways from searches of complex electronic devices under the Fourth Amendment" under the Supreme Court's new decision in Riley v. California. After discussing Riley at length, Judge Donald writes:

The likelihood that an electronic device will contain 1) many kinds of data, 2) in vast amounts, and 3) corresponding to a long swath of time, convinced the Riley Court that officers must obtain a warrant before searching such a device incident to arrest.

We reach the same conclusion regarding the private search doctrine in the case at bar. As with any Fourth Amendment inquiry, we must weigh the government's interest in conducting the search of Lichtenberger's property against his privacy interest in that property. That the item in question is an electronic device does not change the fundamentals of this inquiry. But under Riley, the nature of the electronic device greatly increases the potential privacy interests at stake, adding weight to one side of the scale while the other remains the same.

This shift manifests in Jacobsen's "virtual certainty" requirement. For the review of Lichtenberger's laptop to be permissible, Jacobsen instructs us that Officer Huston's search had to stay within the scope of Holmes' initial private search. 466 U.S. at 119. To accomplish this, Officer Huston had to proceed with "virtual certainty" that the "inspection of the [laptop] and its contents would not tell [him] anything more than he already had been told [by Holmes.]" Id. That plainly was not the case. As the district court found, "there was absolutely no virtual certainty that the search of Lichtenberger's laptop would have" revealed only what Officer Huston had already been told. Lichtenberger, 19 F. Supp. 3d at 759; see also id. ("[T]he search of a laptop is far more intrusive than the search of a container because the two objects are not alike. . . . [G]iven the amount of data a laptop can hold, there was absolutely no virtual certainty" as there was in Jacobsen.).

Considering the extent of information that can be stored on a laptop computer-a device with even greater capacity than the cell phones at issue in Riley-the "virtual certainty" threshold in Jacobsen requires more than was present here. When Office Huston arrived, he asked Holmes to show him what she had found. While the government emphasizes that she showed Officer Huston only a handful of photographs, Holmes admitted during testimony that she could not recall if these were among the same photographs she had seen earlier because there were hundreds of photographs in the folders she had accessed. And Officer Holmes himself admitted that he may have asked Holmes to open files other than those she had previously opened. As a result, not only was there no virtual certainty that Officer Huston's review was limited to the photographs from Holmes's earlier search, there was a very real possibility Officer Huston exceeded the scope of Holmes's search and that he could have discovered something else on Lichtenberger's laptop that was private, legal, and unrelated to the allegations prompting the search-precisely the sort of discovery the Jacobsen Court sought to avoid in articulating its beyond-the-scope test.

All the photographs Holmes showed Officer Huston contained images of child pornography, but there was no virtual certainty that would be the case. The same folders- labeled with numbers, not words-could have contained, for example, explicit photos of Lichtenberger himself: legal, unrelated to the crime alleged, and the most private sort of images. Other documents, such as bank statements or personal communications, could also have been discovered among the photographs. So, too, could internet search histories containing anything from Lichtenberger's medical history to his choice of restaurant. The reality of modern data storage is that the possibilities are expansive.

The court holds:

In light of the information available at the time the search was conducted, the strong privacy interests at stake, and the absence of a threat to government interests, we conclude that Officer Huston's warrantless review of Lichtenberger's laptop exceeded the scope of the private search Holmes had conducted earlier that day, and therefore violated Lichtenberger's Fourth Amendment rights to be free from an unreasonable search and seizure. The laptop evidence and evidence obtained pursuant to the warrant issued on the basis of its contents must be suppressed.

I see this as adopting a data-based or maybe file-based approach instead of the physical-computer based approach adopted by the Fifth and Seventh Ciruits. I think the Sixth Circuit's data-based approach is correct. Here's what I wrote in my article on this issue:

Which is better: the virtual file approach. . . or the physical storage device approach . . . ? In my view, the virtual file approach is clearly preferable. Computers are searched to collect the information they contain. When assessing how the Fourth Amendment applies to the collection of information, courts should focus on that information rather than the physical storage device that happens to contain it. Using the physical box as the common denominator of a computer search would also lead to unpredictable, unstable, and even disturbing results. As computers contain more and more information over time, it becomes increasingly awkward to say that a second search through the contents of a computer simply examines the contents of the physical box in a more comprehensive manner than before.

This point is all the more clear in a networked world. A single physical storage device can store the private files of thousands of different users. It would be quite odd if looking at one file on a server meant that the entire server had been searched, and that the police could then analyze everything on the server, perhaps belonging to thousands of different people, without any restriction. Furthermore, a single file on a network may actually be stored in several physical boxes. Some computer storage devices may not be stored in any boxes at all. Over time, it should become increasingly clear that the Fourth Amendment should track the information, not the physical box.

Having rejected the physical storage device as the proper zone of a computer search, the next question is a subtle one: is the proper zone the virtual file or the exposed data? Existing cases tend to ignore this question because they mostly involve possession of digital images of child pornography, in which the contraband image is both the file contents and the exposed data. The distinction between files and data collapses in this context though in other cases the distinction will prove tremendously important. Imagine that an officer executing a search warrant comes across a computer that is up and running with the first page of a one hundred page document on the screen. The officer wants to view the other ninety-nine pages of the document to see if they reveal evidence of criminal activity. If for some reason the officer cannot justify a "search" of the computer, can the officer take the mouse and scroll down to read the rest of the one hundred page file without conducting a search? Or does publishing the rest of the document on the screen search that information?

I think the better answer is to use the exposed information as the common denominator. The scope of a computer search should be whatever information appears on the output device, whether that output device is a screen, printer, or something else. Under this approach, scrolling down a word processing file to see parts of the file that were previously hidden is a distinct search of the rest of the file. This approach works best for several reasons. First, much information stored on a computer does not appear in a file.108 If the law is keyed to files, how can it apply to information not stored in a file? Second, this approach fits nicely with the exposure-based approach to searches. Once again, what matters is exposure to human observation. Third, virtual files are not robust concepts. Files are contingent creations assembled by operating systems and software. Fourth, an analyst who takes a mouse, clicks, and pulls down the file to see parts of the file not previously exposed has done nothing different from another analyst who double clicks on a second file to open it. In both cases, the analysts are exposing information not previously exposed. Both actions should be treated as searches.

Four additional thoughts, for the loyal readers (anyone?) still with me.

First, in her new opinion, Judge Donald argues that the Fifth and Seventh Circuit decisions are consistent with her decision. In those cases, she notes, the courts were applying the same tests. But although those courts applied the same tests, they used different units. And the unit you choose determines the result you get. The Fifth and Seventh Circuits have taken the view that when a single file has been searched by a private party, the entire physical device has been searched and the government can search the entire computer without a warrant. Under that approach, this case would have been an easy win for the government. Because Holmes had searched some files on the computer, the government could search the entire computer. So the unit you pick determines the outcome.

Second, there's a wrinkle about the case that seems worth flagging. The officer presumably wanted Holmes to reconstruct the private search. That is, he probably wanted her to open the same files as she had seen. It's not clear from the opinion exactly what words he told her, but that seems the likely intent. Given that, does it matter that the opening of new files was her act and not the officer's? For example, what would the result be if the officer had said to Holmes, "I want you to reconstruct the exact search you conducted before, only opening the exact files that you have previously opened." What happens if Holmes does her best but ends up opening different files? What if Holmes just ignores the officer at that point and browses through the computer? Would that be a private search by Holmes? Or does that become a government search because the officer directed it, even if he wanted it to be different from what it was? I'm not sure.

Third, I would think the issue of whether the files opened had been opened before would be a matter of fact-finding for the district judge. The court notes the uncertainty on that, but I would think that's just an issue for below, with appropriate reference to the burden of proof.

Fourth, you might be asking, can't the government just get a warrant based on the private party report? It depends. Sometimes the private report will provide probable cause and sometimes it won't. The officers may want to check to make sure there is probable cause, making reconstruction of what the private party saw helpful to getting a later warrant. The question is when and how they can do that.

As always, stay tuned.