This past July, the National Highway Traffic Safety Administration warned autonomous vehicle developers about “a clear pattern” of driverless cars interfering with police officers, firefighters and other first responders, urging companies to address the problem immediately. The warning comes after a growing series of real-world incidents that raise serious questions about the tendency for these vehicles to make complex emergency situations worse.
From my vantage point as the former New York City fire commissioner, I am terrified of the chaotic public safety disaster that awaits when driverless AI robocars flood our streets with little to no way to control them. Given the risks, I strongly urge Gov. Hochul and the Legislature to put the brakes on allowing driverless AI robocars to take over New York.
Examples of AI robocars directly threatening public safety and impeding first responders are abundant. Last July, as fireworks filled the skies of San Francisco, multiple Waymo vehicles drove directly into areas where fireworks were actively being set off. In one widely shared video, a driverless car rolled over a lit firework in the middle of the street while passengers remained inside. In another incident, a Waymo vehicle was struck by fireworks and caught fire, while other Waymo vehicles became stranded in gridlock as crowds filled the streets.
Unfortunately, these weren’t isolated incidents. Earlier this year in Los Angeles, a Waymo vehicle carrying passengers drove into an active police scene. In Austin, emergency responders reported that a Waymo vehicle blocked an ambulance responding to a mass shooting. During severe flooding in Atlanta, Waymo suspended service after its vehicles struggled to safely navigate flooded streets.
These incidents happened hundreds of miles apart under completely different circumstances, yet they point to the same concern: autonomous vehicles continue to struggle when the environment becomes unpredictable.
Every emergency is different. After decades responding to fires, medical emergencies and disasters across New York City, I have learned that no emergency scene unfolds according to a script. Police establish perimeters. Firefighters stretch hose lines across intersections. EMTs wave vehicles through openings that didn’t exist 30 seconds earlier. Decisions are made in real time, often communicated with nothing more than eye contact, hand signals and instinct built from experience.
Supporters of autonomous vehicles often point to how these systems perform during routine trips. My career taught me to ask a different question: what happens when the routine disappears? Public safety isn’t measured on an ordinary Tuesday afternoon. It’s measured during emergencies because those are the moments when first responders need everyone on the road to make good decisions.
New York presents challenges unlike almost anywhere else in the country. Our streets are narrower, our traffic is denser and our emergency scenes are more complex. Professional drivers navigate those changing conditions every day by communicating with first responders and responding to directions that can’t be anticipated by software.
New Yorkers have always embraced innovation, but we’ve also understood that public safety cannot be treated as an experiment. The real-world incidents we’ve already seen across the country should give every policymaker pause before allowing autonomous vehicles onto New York’s streets.
As a fire commissioner, I spent my career preparing and investigating the moments when everything goes wrong. AI robocars may perform well under ordinary conditions, but emergencies are anything but ordinary. They demand judgment, communication and the ability to recognize that every situation is unique. The incidents we’ve already seen around the country make one thing clear: driverless cars are not ready for New York.
Cassano is a former commissioner of the New York City Fire Department.