The use of artificial intelligence (AI) in technology, law enforcement, and courts raises complex and unprecedented questions about accountability, the application of “human” rights to autonomous intelligences, and how the Constitution interacts with nonhuman intelligences. Courts must consider how “human” rights, duties, and laws should apply to these machines, especially in the context of the Constitution.

Intuitively, AI tools can engage in massive searches and present results that are difficult or impossible to explain, even if the outputs seem correct. Use of a facial recognition AI for identification of a defendant may form the basis for law enforcement action against a specific individual without any human intervention or scrutiny of that decision. These AIs have led to wrongful arrests and rescinded warrants and have been found to increase racial disparities in arrests. The AI cannot be put on the stand to defend its selection of that person, and the law enforcement officers who use it cannot thoroughly explain how it makes its decisions. This is no hypothetical scenario: Clearview AI has been used nearly a million times for facial recognition purposes by U.S. police.