Navigating the future: Creating a compensation framework for autonomous vehicle incidents

Introduction

Driverless cars and other autonomous vehicles (AVs) present real opportunities to increase road safety and decrease emissions, along with other substantial benefits to the car industry and passengers. But one glaring question surrounds this promising development: Are autonomous vehicles safe enough? There is a critical, lurking regulatory and legal issue flowing from that question. Federal regulators must address how to promote—or at least not impede—this new technology that will likely save tens of thousands of lives per year but will nonetheless cause deaths and injuries. And industry participants must grapple with whether potential liability from those deaths and injuries pose too great a threat to invest resources in the first place. 

We propose a victim compensation fund established by Congress to provide an efficient way for victims in autonomous vehicle accidents to bypass the court system. Although some have previously noted such a possibility, the issue is more important now than ever in light of technological progress and the current political environment. The new administration has taken a keen interest in promoting fully autonomous vehicles, and there may be greater appetite in Congress to assist the administration’s efforts. A compensation fund could also ease public concerns about whether consumers should trust the idea of widespread autonomous vehicle adoption and use.

Potential benefits of fully autonomous vehicles

The promise of autonomous vehicles extends far beyond convenience, harnessing the power of artificial intelligence to offer substantial societal benefits across multiple dimensions described below 

Safety

Around 40,000 people in the United States die each year in traffic fatalities. That is nearly the population of Burlington, Vermont. Human error contributes to approximately 94% of serious crashes, suggesting enormous potential for improvement through automation. AVs can potentially eliminate many of these errors through:

• elimination of drunk, distracted, and drowsy driving. Drunk driving alone causes over 13,000 fatalities annually;
• consistent adherence to traffic laws and speed limits;
• use of multi-sensor systems with 360-degree awareness that can detect objects better than humans;
• reaction times that are much faster than humans;
• technological ability to communicate with other vehicles and infrastructure.

The net effect of these improvements seems very likely to be a significant drop in automobile fatalities. These safety improvements will also generate significant economic benefits by potentially reducing health care costs and decreasing insurance rates.

None of this is to imply that autonomous vehicles are free of safety issues. For example, valid concerns remain—the complicated sensor and computing technology could malfunction, causing an accident. This may be especially likely under certain driving conditions such as dawn or dusk. Further, the algorithm guiding autonomous vehicles may be faulty or may be hacked. Finally, human drivers may create accidents by errantly overriding the AV.

Environmental

Vehicles are a significant source of pollution, impacting humans and the environment. The typical car emits over four metric tons of carbon dioxide every year. Autonomous vehicles are more fuel efficient and therefore less harmful to the environment because they drive more efficiently than drivers who idle, brake inconsistently, and do not use consistent speeds. One estimate suggests that smoother traffic flow and optimized acceleration and deceleration patterns may reduce fuel consumption by 18% and carbon dioxide by 25% compared to human drivers. Beyond the environmental benefits, smoother traffic flows should decrease traffic congestion, saving time and increasing productivity.

Increased mobility

Autonomous vehicles can greatly increase mobility for the millions of Americans who cannot drive, including seniors and those with travel-limiting disabilities. This will be all the more important as the United States population ages in coming years. Many of people who cannot drive also lack ready access to public transportation, complicating their ability to commute to work and access vital services. Increasing mobility for these individuals will have significant economic benefits by increasing the workforce.

The (absence of a federal) regulatory landscape

Many states regulate the development of autonomous vehicles, and the federal government has similarly been examining the extent to which these vehicles can operate safely. But currently there are no federal statutes or legally binding rules that affirmatively allow autonomous vehicles without human drivers to self-drive on roads. As the technology develops, however, Congress and federal agencies will be pushed to regulate in this space.

A central impediment to AV deployment is the uncertainty surrounding liability when accidents inevitably occur. Current legal frameworks are ill-suited to autonomous vehicle technology. Fault-based insurance frameworks generally assume human driver control and responsibility. With respect to tort law, key issues such as the chain of causation and resulting legal liability can become convoluted when a product involves multiple software, hardware, and service providers as well as the human driver. More fundamentally, determining whether an autonomous system operated “reasonably” in the eyes of the law is an open challenge, and courts will struggle with deciding whether an algorithm acted negligently in an accident. This mismatch between existing legal frameworks and new technology creates significant uncertainty for courts, manufacturers, drivers, and insurers, potentially leading to inconsistent outcomes and unpredictable liability exposure.

This uncertainty over legal liability is likely to deter development and deployment of autonomous vehicles. Insurers may refuse to issue policies given challenges in estimating liability and setting premiums. Manufacturers may also be deterred by this uncertainty. This may be especially true for smaller providers, which are important in spurring innovation. The upshot could be a paradoxical scenario where a technology that could save tens of thousands of lives annually faces barriers due to concerns about a smaller number of inevitable incidents.

The proposal: A victim compensation fund

“Victim compensation funds” are used to help those who are those affected by national disasters receive monetary payments, usually in response to potential litigation. After the September 11 terrorist attacks, for example, Congress established a fund for victims who waived their right to sue in court. These funds are not always run exclusively by the federal government. After the Deepwater Horizon oil spill, private entities agreed to a multibillion trust fund to pay claims resulting from the spill.

Although victim compensation funds typically respond to discrete events like the September 11 attacks, there is no inherent reason why they can only be used in that scenario. And there is historical precedent for compensating victims outside of a courtroom. The Vaccine Injury Compensation Program provides money to eligible individuals injured by standard childhood vaccines. We propose that Congress and leaders in the autonomous vehicles industry explore establishing a victim compensation fund for those injured by autonomous vehicles. 

The benefits of such a fund for those injured or killed could mitigate a major impediment to the development of driverless cars: the potential for enormous liability in tort for victims of autonomous vehicles accidents, even where driverless cars save thousands of lives overall. To use a hypothetical example, if autonomous vehicles cut the number of annual traffic fatalities in half but cause a much smaller number of deaths and injuries, it would be much more beneficial to transition to driverless cars. A fund that compensates victims for death and injury may smooth the transition to autonomous vehicles while providing greater incentives for industry to invest in autonomous technology necessary to make the potential a reality.

We recognize that an autonomous-vehicle victim compensation fund is not a comprehensive solution to every problem raised by the technology. The fund would operate as a substitute for the tort system only. Other accountability mechanisms like criminal or civil investigations would still be available where industry participants’ wrongful conduct extends beyond an individual accident. As a substitute for the tort system, the fund will channel claims away from the court system, which may slow the judiciary from defining reasonable standards of care to adjudicate claims for victims that do not participate in the fund because it will have fewer opportunities to do so. But not everyone will participate, so there will still be a role for courts to play.

Our proposal is just a sketch, and Congress and federal regulators will need to confront several questions if the federal government pursues this option. Among them:

• Administration. Who should administer the fund? We find compelling a proposal to entrust the fund to the National Highway Transportation Safety Institute given that agency’s subject matter expertise and ability to solicit public input but note that courts may be more independent administrators.
• Regulatory framework. Can the fund be established before the federal government sets up a regulatory framework for autonomous vehicles to operate on our nation’s roads? How will the fund align with forthcoming federal regulations on AV operation and certification? Ideally, the fund would work as part of a comprehensive federal regulatory framework that includes autonomous vehicle safety standards, testing protocols, and certification requirements.
• Funding. How will contributions from manufacturers who choose to participate in the fund be assessed? One possibility is a contribution formula based on the number of miles driven by each manufacturer’s automated vehicles. What portion of dollars will come from the federal government, which typically provides the funds?
• Participation eligibility. If the federal government establishes a regulatory framework, will affirmative federal approval to operate autonomous vehicles be necessary for a manufacturer to benefit from the fund?
• Manufacturer eligibility. Should manufacturers be excluded from participating if they acted negligently or recklessly? At a minimum, the fund should include provisions to exclude manufacturers who deliberately violate safety standards or engage in fraudulent practices. At the same time, it seems that the fund should cover incidents resulting from good-faith design decisions or unforeseen circumstances.
• Victim fault. What role, if any, will potential victim fault play in the provision of compensation?
• Provision of data. What accident and safety data must manufacturers provide as a condition of participating in the fund? Which of this data would be made public?
• Transition period. How will the fund address the transition period of mixed autonomous and human-driven vehicles that poses unique risk profile?
• Payment determination. How will the fund set payments for standardized injuries? How will non-economic losses such as pain and suffering or emotional distress be addressed?  Will there be an individualized process to consider severe injuries?

As these types of questions demonstrate, this type of proposal is still in its infancy. The time to think about these questions, however, is now. The future of safe transportation may depend on it.

Conclusions

Autonomous vehicles present one of the most promising opportunities to dramatically improve transportation safety while delivering significant environmental, economic, and social benefits. However, realizing this potential requires addressing the fundamental liability challenge that could otherwise impede development and deployment. The proposed victim compensation fund offers a balanced approach that would protect innovation while ensuring appropriate compensation for those harmed. This framework could help resolve one of the most significant barriers to widespread adoption of autonomous vehicles.

By acting now to establish a comprehensive approach to autonomous vehicle regulation, including a specialized compensation mechanism, policymakers can help ensure that the life-saving potential of this technology is realized while protecting those who may be harmed, particularly during the inevitable transition period.