Last summer, in United States v. Ganias, a Second Circuit panel held that the government violated the Fourth Amendment when, for two-and-a-half years, it retained digital copies of files that were outside the scope of an initial warrant and then used those files in a subsequent investigation. Applying the exclusionary rule, the Second Circuit panel vacated the defendant’s conviction. The government petitioned for panel rehearing on the application of the exclusionary rule, without contesting the holding that the Fourth Amendment precluded the lengthy retention of digital copies of files without a warrant.
Just last week, however, the Second Circuit sua sponte granted an en banc rehearing on not only the application of the exclusionary rule but also on the Fourth Amendment holding. The original Ganias decision had broad implications for the application of the Fourth Amendment in the digital age—it was the first court of appeals decision to directly address the retention of digital duplications. Its efforts to “adapt traditional Fourth Amendment concepts to the Government’s modern, more sophisticated investigative tools” are surely needed. The rehearing en banc on the Fourth Amendment issue as well as the exclusionary rule issue is thus quite significant.
The basic facts of Ganias are not unique. In the course of a fraud investigation in 2003, government investigators obtained and executed a warrant for certain computer-stored data. To facilitate the sorting of that data, the government made a mirror image of the computer hard drives for off-site review instead of seizing the physical computers. Some time later—two-and-a-half years later—the government obtained a second warrant to search the same files, which it already had in its physical possession, for evidence of additional crimes by a different individual. The defendant claimed that this second search—even though pursuant to a warrant—violated his Fourth Amendment right to be free from unreasonable searches and seizures.
The Second Circuit panel agreed: The 2003 “warrant for the seizure of particular data on a computer” did not permit the government “to seize and indefinitely retain every file on that computer for use in future criminal investigations.” The court focused on two facts: first, the initial warrant pertained to a different investigation of different individuals, and second, the gap in time between the two warrants. The court also was not convinced by the government’s argument that it needed to retain the data to authenticate evidence in the first investigation.
The decision was significant because it characterized the retention of lawfully obtained data as an unreasonable seizure. But, as other commentators have recognized, the Ganias decision created as many questions as it answered. The court acknowledged that the government could reasonably seize the data in the first place—so at what point did the retention become unreasonable? Would it have been unreasonable to expand the search to a separate defendant if done immediately after the initial warrant was executed? Would it have been unreasonable to expand the search two-and-a-half years later if it was done for the same investigation? Perhaps the en banc court intends to address these concerns.
The panel characterized the retention as an unreasonable seizure, but implied that it only became unreasonable when the government used the files for the second investigation. If we conceptualize seizures as invasions of possessory interests—as the Ganias decision did, calling it an interference with the right of exclusive possession—then the reasonableness or unreasonableness shouldn’t turn on what the government does with the seized property. That is, in many senses a seizure is binary—either possessory interests are infringed or they are not—and a seizure’s reasonableness is usually measured by the length of time of the seizure. In Ganias, though, the panel measured reasonableness not only by time but also by what the government did with the property.
An alternative approach would be to conceive of retention of digital duplications as ongoing searches, rather than seizures. After all, the concern with continued retention of electronic copies is centered on privacy interests, not possessory interests. Thus, conceiving of data retention as a search fits more naturally into the traditional possessory/privacy dichotomy. Indeed, the courts that have considered duplications, including the Ganias court, have focused on the privacy interests at stake, not the possessory interests.
Viewing data retention as a search would also have several other implications. For one, it would allow the court to focus the reasonableness inquiry on how the government used the data. While retention alone might infringe on reasonable expectations of privacy, such an infringement solely for the purpose of data authentication would not be unreasonable, and thus would not violate the Fourth Amendment. Second, if data retention is a seizure because it interferes with the right to exclusive possession, then an individual could assert a Fourth Amendment claim against government retention of publicly shared data. A defendant might post violent threats on Facebook, for example, but retain “ownership” of their content and thus could assert his right to exclude others from viewing that data—including the government if it obtained copies of the threats without a warrant. But if data retention is a search, then the poster has would have waived any claim of privacy once he shared the data with the world.
Several commentators have recognized the peculiarity of calling duplication or retention a search if the government isn’t actually looking at the contents. But it is also odd to characterize it as a seizure, given that the individual retains the original. When the Second Circuit grapples with these issues anew in September, it may also consider the impact of its ruling on other data retention programs—especially if Justice Sotomayor convinces her colleagues that the third party doctrine, under which an individual has no reasonable expectation of privacy in data shared with a third party, should be curtailed.