In a recent article by Forbes, attention was brought to the case of David Zayas who was arrested in Westchester County, New York. He later pled guilty to a drug trafficking charge. This arrest originated after an Artificial Intelligence (AI) program, having analyzed extensive vehicle data captured by Automatic License Plate Recognition (ALPR) software, determined that Zayas’ car was on a journey typical of a drug dealer. As a result, Westchester Police Department pulled the vehicle over and searched it. They found 112 grams of crack cocaine, a firearm and $34,000 in cash.
This raised constitutional question for me, how does the development and such use of this technology infringe on our 4th amendment rights? The case that immediately sprung to mind was Carpenter vs United States which addresses the question of whether the warrantless search and seizure of cell phone data, specifically cell site analysis in tracking a persons whereabouts, violates the Fourth Amendment.
The Supreme Court concluded, in a 5-4 decision that the warrantless search of Carpenter’s cell phone records was unconstitutional. So, how does this apply in the case of Zayas and is there an argument here based on the use of the AI and ALPR technology?
The majority opinion in Carpenter opens up referencing the history of the Fourth Amendment and states, “The “basic purpose of this Amendment,” our cases have recognized, “is to safeguard the privacy and security of individuals against arbitrary invasions by governmental officials.”” It specifically references the founders intent to prevent the unrestrained search for evidence of criminal activity conducted by British officers during the colonial era. Also pointing out that this subject was a catalyst to the Revolution itself.
However, there are also some considerations in the case law about advancements in technology and reasonable expectations of privacy when traveling on a public thoroughfare. US v Knotts references that a person does not have a reasonable expectation of privacy here, because their movements and location were voluntarily conveyed to anyone who wanted to look. The Knotts opinion did state that “different constitutional principles may apply if twenty four hour surveillance of any citizen of this country were possible.” It should also be noted that Knotts was in 1983, some 40 years ago when technology such as AI was only in science fiction movies.
The Supreme Court in Carpenter also point to Katz and Jones for reference. Stating that privacy, even in an area accessible to the public, may be constitutionally protected and that a person has a reasonable expectation of privacy in the whole of their physical movements.
While addressing the data collection itself it is noted that this case is unlike Jones, where the subject of GPS tracking was “targeted”. This particular method of data collection is continually logged, for everyone, not just those under investigation and “police need not even know in advance whether they want to follow a particular individual, or when.”
This particular point is very much in line with the Zayas matter. To my knowledge, Zayas had not been developed as a suspect beyond this autonomous data collection and AI analysis. Traveling routes on public thoroughfares is not illegal. Drug traffickers also use the road system. Should that correlation open all people up to suspicion? In the Zayas case the AI was correct, but how many innocent civilians have been stopped by the government based on this data alone? We will likely never know.
AI technology is not new, but it is becoming much more prevalent in our digital world. Its capability to analyze and interpret large datasets is impressive and vastly exceeds the human capability on large data. Because of this it has a wide range of extremely useful applications. However, as the use of AI expands into the realm of investigations, and particular government surveillance, it may be time for some safeguards. I would be curious to see its use tested in the Supreme Court. For now though, the opinions of our members would suffice.