Self-driving technology has moved from experimental labs to busy intersections in cities like Phoenix, San Francisco, and Los Angeles. With companies like Waymo and Cruise expanding their fleets of autonomous taxis, a new era of transportation is officially underway. But even as the technology evolves, one question keeps resurfacing. Are human safety drivers inside these vehicles enough to prevent accidents?
In the early days of autonomous vehicle testing, safety drivers were required to sit behind the wheel to take control if something went wrong. Even today, many cities and states still mandate their presence, especially during pilot phases. The role sounds straightforward, but it’s deceptively complex.
Safety drivers must remain fully alert during a drive that is often uneventful. It’s human nature to tune out during repetitive or passive tasks. And that’s where the real danger lies.
A widely cited study by the National Highway Traffic Safety Administration (NHTSA) warns that passive monitoring of automation often leads to slower reaction times. Unlike a typical driver who is fully engaged in operating a vehicle, safety drivers may struggle to respond quickly if something unexpected happens.
Autonomous taxis rely on an array of cameras, radar, lidar, and machine learning algorithms to detect and respond to their surroundings. In ideal conditions, these systems can outperform human drivers in many areas: reaction time, 360-degree awareness, and strict rule adherence, for example.
But what about unpredictable human behavior? A pedestrian darting across traffic, a cyclist weaving between cars, or a driver ignoring a stop sign are situations that even well-trained humans find challenging. Machines have limitations when it comes to interpreting intent or making judgment calls that rely on context.
That’s where safety drivers are supposed to step in. But again, if they’re not fully engaged every second of the trip, their ability to take meaningful action is in question.
Companies are taking steps to improve the role of safety drivers. Some now require them to complete additional training programs focused on monitoring automation and handling emergencies. Others rotate drivers more frequently to prevent fatigue and complacency.
Some industry leaders are also investing in enhanced monitoring systems that can detect whether a safety driver is paying attention, similar to systems in newer consumer vehicles that issue alerts when a driver appears distracted or drowsy.
Still, critics argue that if a system isn’t safe without a human ready to intervene, then it isn’t truly autonomous.
For autonomous taxis to gain widespread public acceptance, safety remains the deciding factor. Riders need to believe the vehicle will respond to every scenario just as well, or better, than a human would.
Some cities are now reviewing policies that determine when and where autonomous taxis can operate. Others are requiring more transparency from companies, including sharing crash data and software updates.
Meanwhile, federal regulators are continuing to evaluate long-term frameworks. The NHTSA has opened investigations into several autonomous vehicle incidents, and future rules may focus more on how to manage the gray area between full autonomy and human backup.
According to Attorney Stephen Bridger from Phoenix, Arizona, “Technology has limits. The real test for autonomous taxis will be whether we’ve done enough to anticipate unpredictable human behavior, both from passengers and other drivers. Without that, safety remains uncertain.”
Stephen Bridger is a lawyer based in Phoenix, currently employed at the Pinal County Prosecutor’s Office. As a state prosecutor, he does not offer legal advice or represent individuals, but works on behalf of the state to hold offenders accountable and promote public safety.
