Beyond George Jetson: Commuting in a Self-Driving Car?

The internet abounds with stories on modern technological marvels that were once the stuff of science fiction. Drones, video phones, robotic personal assistants, and even treadmills for pets, for example, each made an appearance in the early 1960s through the animated series The Jetsons. Despite all the futuristic conveniences in George Jetson’s fictional life, however, he was not exempt from the workaday task of driving himself to the office (albeit in a bubble-topped jet). But today, for those who have resumed going to the office in the wake of COVID-19, we may be close to relief from the daily rush-hour commute, as self-driving cars—officially known as “autonomous vehicles”—continue to advance in reliability, popular acceptance, and regulatory authorization.

As with any new technology, the introduction of self-driving cars on public roadways raises multiple questions for both policymakers and ordinary citizens. To what extent can a car be “self-driving”? And if a car really does drive itself, who should be responsible when things go wrong? And at what level of government should these questions be addressed? Thus far, questions of liability and regulation of autonomous vehicles have largely been handled at the state level. But as such vehicles continue to grow in use and popularity, scholars and policymakers have begun to question whether ordinary tort law provides sufficient regulation or whether a new and uniform set of standards is necessary.

How Did We Get Self-Driving Cars?

While the idea of a self-driving car is almost as old as the automobile itself, it was not until the 1980s that engineers first made significant strides toward bringing the idea into reality. Aided by funding from the Defense Advanced Research Projects Agency (DARPA), university engineering departments began to incorporate sensor technology and artificial intelligence software into new prototypes, enabling them to navigate city streets like a human driver.

One such prototype, Carnegie Mellon’s “NavLab 5,” achieved a major milestone in 1995 when it completed a nearly 3,000-mile drive across the United States, covering all but 50 miles in autonomous mode. DARPA sought to build on this momentum with “Grand Challenges” in 2004 and 2005, offering seven-figure cash prizes for autonomous vehicles that could conquer difficult desert routes. Although no one succeeded in the original contest, a Stanford team won the second one (and $2 million) in October 2005 when its entry successfully negotiated the designated 132-mile off-road course in under seven hours.

A little more than a decade later, the technology had advanced well beyond prototypes into commercial use. In December 2018, Waymo announced that it would allow pedestrians in the Phoenix suburb of Chandler, Arizona, to order self-driving taxis. Although Waymo’s service was limited geographically, the idea gained expanded reach three years later when Cruise began offering a similar service in San Francisco. Comparable driverless taxi services have been introduced in specified urban areas overseas as well, including China and Japan.

Despite the undeniable technological advances of such offerings, analysts have questioned whether the cars involved are appropriately deemed “self-driving,” given that they are usually monitored remotely and subject to intervention by human engineers when needed. For this reason, the Society of Automotive Engineers (SAE) has advanced the view that the term “automated vehicle” may be applied in varying degrees along a continuum of capabilities, with commonplace features such as cruise control and crash-avoidance warnings at one end (Level One) and fully autonomous, no-intervention, “steering wheel optional” operation at the other (Level Five). The National Highway Traffic Safety Administration has adopted SAE’s classification system, such that it is now U.S. Department of Transportation policy.

What Laws Govern Autonomous Vehicle Technology?

Since 1966, automotive safety standards have been set at the federal level through the Federal Motor Vehicle Safety Standards (FMVSS). But the FMVSS are principally mechanical in nature, focused on automotive design, leaving issues of driver licensing, vehicular operation, and ordinary traffic laws to the states. In consequence, there is no uniform federal law in the United States governing the operation of self-driving cars. Instead, the National Highway Traffic Safety Administration has issued highly flexible “guidance” for states to consider as they craft their own laws and regulations to maintain public safety while encouraging innovation.

States have rapidly risen to the task. As of 2015, only California, Florida, Michigan, Nevada, Virginia, and the District of Columbia had enacted legislation governing the testing of self-driving cars on public roadways. Two years later, the number of states that had introduced such legislation had risen to 28. Today, 40 states plus the District of Columbia have passed laws relating to the testing, operation, and even commercial deployment of cars of varying degrees of automated control.

Unsurprisingly, leaving states to their own devices has led to a variety of laws across jurisdictions. Some states, such as Arizona and California, allow the commercial operation of for-profit driverless transportation services, subject (in some instances) to certain minimum requirements for liability insurance. Others, such as Connecticut and Massachusetts, do not permit the full deployment of automated vehicles on public roads but will allow the testing of such vehicles in autonomous mode, provided that a licensed operator remains in the vehicle at all times. And still others, such as Idaho and Oregon, appear not to have enacted any specific legislation on the issue at all.

What Is Next for Self-Driving Cars?

The current lack of uniformity among state laws on autonomous vehicles has prompted concerns about stifling innovation, and Congress has responded. In September 2017, separate bills were introduced in both the House and Senate that would have enabled federal preemption of any state law regarding autonomous vehicle design. Although neither bill became law, Congress is expected to take up the issue anew now that the November 2022 election has passed.

Critics have questioned, however, whether design preemption goes far enough to encourage innovation. Among other things, they have noted that traffic laws vary widely from state to state, and thus, to operate nationwide, a vehicle’s operating software must be programmed with the laws of 50 different states. And this is to say nothing of the significant variation in tort laws in the different states, which could make operation in certain jurisdictions significantly riskier from a liability perspective. Addressing concerns of this nature would, of course, require federal preemption not merely of laws regarding self-driving cars, but of traffic and tort law in general.

At the other end of the spectrum, consumer advocates have questioned whether the rush to develop new technology has overcome legitimate safety concerns, noting that robotics is a very new field, the full hazards of which are unknown. But technology advocates enjoy pointing out that human error remains by far the most prevalent cause of vehicular injuries and that unfounded fears accompanied the launch of many ordinary machines—elevators, which were once driven by human operators, being a prime example—whose automation we now take for granted.