A public transit system can be run as expected and still not be trustworthy. As artificial intelligence (AI) enters maintenance, operations, and vendor-supplied technologies, transit leaders face a consequential question: what evidence shows that the systems supporting safe, reliable service are operating as intended?
