5 minute read

Where Autonomous Vehicles and the Law

Where Autonomous Vehicles and the Law Collide

Abinaja Yogarajah

Advertisement

Autonomous Vehicles Whilst the unchartered territory of Autonomous Vehicles (‘AV’) is quickly approaching us, Australian and international legal structures may not be able to support the novel situations that accompany it. AV are self-driving vehicles that are capable of sensing their environments and moving safely with little to no human input. The insertion of Artifcial Intelligence (‘AI’) at every stage of AV operation creates many questions where the answers may not yet exist. There are many problems that need to be discussed before AV enter the midst of our inquires, such as patenting, data security, privacy, moral concerns and criminal culpability. Whilst this article may not answer all your questions, it will certainly create some.

Patenting in the Autonomous Vehicle Era AV are dependent on hundreds of complex systems with many components interacting interdependently with each other. They vary from range-sensing and GPS systems to imaging sensors and so on. These outputs are then processed on the AV’s computing system to fulfl the motion objectives of the vehicle in a safe and efcient manner. With so many diferent types of systems and automations involved in the functioning of a car, it is unavoidable that various companies have to collaborate to create functional and dependable driving experiences.

Therefore, AV companies have been relentlessly pursuing the myriad of patenting opportunities provided by complex vehicles. Whilst your standard smartphone has more than 250,000 patents in the US alone, the rapid increase in patenting by the AV industry, combined with the complexity of AV systems, indicate that AV will likely revolutionise intellectual property protection.

One of the biggest concerns in the intellectual property law arena is whether current Australian and international patent regimes are adequate in protecting AI

However, when we are found in moral dilemmas in reality, we very often behave diferently. AV, on the other hand, will behave exactly as it is programmed.”

systems that AV function on. The premise of AI or machine-learning is based on the incremental evolution of underlying systems and algorithms. It strives to duplicate human intelligence by interacting with the world and correcting itself. Given the constantly evolving nature of technology, it raises the difculty of capturing the invention into a narrow description as required for a patent grant.

Who Owns It? Another issue arises where AI’s self-learning process allows subject matters to be created by the AI itself. If we suppose that the fnal output is patentable, then who is the rightful owner? Many jurisdictions such as the US and UK have diferent answers, but there is currently no legal recognition of an AI software program as a legal person, which may change one day.

Avanci Many of the technologies that operate AV are owned by information technology, telecommunication and software companies. As such, many of the patents involved with AV components will acquire a competitive advantage by becoming exclusive and applying restrictions on their licenses. However, these AV technologies relate to how cars communicate and share data, and as a result, they become essential standards for communication, safety and security. A company that seeks to tackle the necessity for diferent AV companies to cooperate and coexist is Avanci. Avanci is an interesting new marketplace that gives users one-stop access to essential patented technology. It was created out of a need for the licenses to essential patents for further research and development. It currently ofers licenses to use essential patented technologies owned by 11 member companies, with others expected to be added.

Data Security and Privacy Concerns As AV technology develops, more safetymonitoring systems are inevitably put in place. AV may collect and maintain identifying information about the owner or passenger for reasons ranging from authenticating authorised use, to customising comfort and entertainment settings. This information is likely capable of identifying owners and passengers with extreme accuracy.

Location data is essential for AV and increasingly will be collected for navigation. It seeks to collect user’s destination information, route information, speed, time travelled, and time spent on the road. Additionally, as explored in United States v Jones, this type of location information ‘generates a precise, comprehensive record of a person’s public movements that refects a wealth of detail about their familial, political, professional, religious and sexual associations.’ As such, cars can be expected to store so much of our personal and sensitive data in the future.

So, what now? Australia’s National Transport Committee (‘NTC’) released a report on AV expressing their concern about its reliance on copious amounts of data to function and the inadequacy of Australia’s current legal structure in protecting individual privacy. They suggest

that whilst the collection of data could enhance government decision-making in law enforcement, trafc management, road safety and infrastructure planning, it must be balanced with privacy protections. During the NTC’s research, many stakeholders recommended that the government should only be allowed to access data for the beneft of the general public. Specifcally, that it should not be used for insurance purposes, commercialisation purposes, surveillance, covert law enforcement or data matching.

Moral Concerns and Criminal Culpability Criminal Culpability What happens in genuinely unforeseeable categories of harm? In the era of AV and multitasking, if a car crash were to happen, how do we attribute liability? What about in cases where a person could have intervened but did not?

Generally, many courts will try to blame it on human error and operation, rather than a machine. Recently, a court attributed an airplane accident to the airline for incorrectly balancing the cargo, regardless of the fact that autopilot was engaged. Having said that, there is a limit as to how much culpability will shift to owners of vehicles, rather than the manufacturer – just for agreeing to their terms and conditions. This was the case in the fatal car crash involving a Tesla AV where the deceased driver was found to have assumed and taken on the risk of engaging autopilot.

Machines and Moral Concerns We have all heard about the ‘Trolley Problem’. There is a runaway trolley barreling down railway tracks. You can either let it kill 5 people or pull a lever which would divert it to kill one instead.

Does AV raise a new Trolley Problem? As AV will have to make quick moral decisions about whom to kill in the event of an accident, it forces us to confront profound philosophical questions.

As humans, we often have theories in our heads and even tell ourselves what the right thing to do is. However, when we are found in moral dilemmas in reality, we very often behave diferently. AV, on the other hand, will behave exactly as it is programmed. Whilst we cannot answer the ethics of machine morality, it is something to consider with the emergence of AV.

Conclusion Emerging technologies, like AV, challenge the adequacies of laws when existing legal categories fails to anticipate its novel materialisations. Due to the nature of machine learning, changes in the law that will be required to accommodate AV will almost always be reactive and unprecedented – for now.

This article is from: