Deprecated: SBM::mongo(): The Mongo class is deprecated, please use the MongoClient class in /sbm/sites/publish/trunk/lib/sbm.php on line 656
University of Maryland professors spearhead new approach to interpreting Fourth Amendment

University of Maryland professors spearhead new approach to interpreting Fourth Amendment

Machine learning technology could provide a solution to clarify search warrants

A big question under debate lately is whether police should be required to get a search warrant before tracking someone’s location. The traditional answer has been “no, you have no expectation of privacy in public movements.” But times are changing.

Recently, researchers at the University of Maryland Francis King Carey School of Law and Columbia Engineering published a study in the New York University Journal of Law & Liberty that examines how advances in machine learning technology may change the way courts treat warrants and privacy issues. Their study focuses on how technology can provide key Fourth Amendment insights, especially when it comes to long-term surveillance and whether a search warrant is necessary.

Today, machine learning is the branch of computer science that studies systems that can draw important insights from data. “We found that machine learning makes it clear that mosaics can be created, and that the duration of investigations is relevant to their substantive Fourth Amendment treatment because duration has a large impact on the accuracy of the predictions,” according to Jebara, who chairs the Institute for Data Sciences and Engineering’s Foundations of Data Sciences Center.

Academics have developed the “Mosaic Theory” of the Fourth Amendment, which holds that a large collection of data is more revealing than the individual points. The Court of Appeals for the DC Circuit accepted this theory in Maynard v. U.S., holding that long-term surveillance was a search protected under the Fourth Amendment, that it exposed an "intimate picture of the subject's life that he expects no one to have" and should have had a warrant.

“The basic idea is very simple: when machine learning techniques can make accurate enough predictions, you have a mosaic,” said Steven M. Bellovin, Computer Science (CS) professor at Columbia Engineering, who co-authored the study.

According to Renée McDonald Hutchins, a professor of Law at University of Maryland Carey School of Law, this has been a controversial subject, especially in this new age of Big Data, and the courts don’t agree about what constitutes a search that requires a warrant.

“It’s controversial because the court has repeatedly said that Fourth Amendment rules should be driven by objective not subjective assessments.  This objective/subjective dichotomy has presented challenges for the Mosaic Theory because what we sense at a gut level -- that knowing a lot of seemingly innocuous data about a person tells you more than just the individual data points -- has been hard to confirm objectively,” she explained in our recent interview.

For example, according to Hutchins, in the Supreme Court's decision in Jones, Justice Sotomayor noted that following someone around for four weeks was too long because it allows you to learn things like whether they go to the psychiatrist, church, a massage parlor or a gay bar. In response, Justice Scalia inquired how we might be able to tell that four weeks was too much time to follow someone without a warrant.  What Scalia was asking for was some objective measure that might help him determine why a constitutional line could be drawn between say four of data and two of data.

“Until very recently, the science did not actually help us answer the questions of line drawing.  But now it is beginning to,” she added.

Now, the legal academics have a better understanding of the value of aggregated data when viewed through machine learning. According to Bellovin, while reasonable minds may dispute the most minimum accuracy threshold, the collection of data allowing predictions that exceed selected thresholds should be deemed unreasonable searches in the absence of a warrant.

According to Hutchins, the solution is not going to come from just the lawyers or just the tech folks. “The answer will only come when the two disciplines begin working together. It’s essential that the legal and computer science communities work together, and that the law on location tracking continues to keep step with the current state of scientific discovery,” she commented.

These days, the law can only regulate technology that enables encroachments on privacy if the law understands what that technology is capable of. The Fourth Amendment is one area of the law that must consistently consider technological enhancements that allow the government to learn more and more about us.  To Hutchins, joining the two disciplines to explore how we might think about problems in the field seemed like a natural development.

“As the two disciplines work more closely, we will be able to conduct the studies that will allow us to make even more targeted recommendations with regard to Fourth Amendment coverage,” Hutchins said.

Over the next decade Hutchins would like to closely study exactly which inferences can be identified with a data set of a particular time length. She added, “As we begin to identify more accurately precisely what it is we can know and how much data we must dump into the machine to learn it, the easier it will be to say with some confidence that a constitutional line has been crossed.”

 For more on fourth amendment issues, check out these articles:

Pivotal Supreme Court case could expand unwarranted cellphone searches

Microsoft GC discusses the future of privacy, Part 1

Law agencies ask wireless carriers for over 1.1 million customer cellphone records

Contributing Author

author image

Amanda Ciccatelli

Amanda G. Ciccatelli is a Contributing Writer for InsideCounsel, where she covers the patent litigation space. Amanda earned a B.A. in Communications and Journalism from...

Bio and more articles

Join the Conversation

Advertisement. Closing in 15 seconds.