In an age of AI and virtual worlds, what responsibility should humans bear?

Existing laws may not be equipped to cope with new technologies such as autonomous driving and virtual worlds. As technologies grow ever-more sophisticated, Professor Tomoyuki Shiraishi of the Faculty of Law explores how human responsibility and duty ought to be regulated.

I specialize in all aspects of civil law. Recently, however, my primary focus has been on laws related to cutting-edge technologies such as autonomous driving, AI, virtual worlds, and avatars.

Existing rules governing liability for damages stipulate that a human must bear responsibility in cases where their actions or judgments are erroneous or negligent. These rules apply specifically to “humans.” However, when an AI acts or makes judgments in an autonomous manner, there is no human involved—in such cases, who should be liable for any ensuing damages?

A prime example is AI-based autonomous driving. Currently, the “vehicle operator”—that is, the person operating and deriving benefit from the vehicle—bears responsibility for any traffic accidents involving that vehicle. In the case of autonomous driving, however, the vehicle operator is not human. Who, then, should bear responsibility: the human sitting in the driving seat, the automobile manufacturer, or the AI itself?

At present, the vehicle operator bears responsibility. However, with fully autonomous driving said to be less than 10 years away, my research explores the applicable legislation, corporate guidelines, and ethical principles.

If harm is suffered in a virtual world, is there a real-world victim?

A researcher must first understand a new technology before he or she can investigate its relationship to laws. When it comes to autonomous driving, for example, I assay to learn from researchers actually involved in developing technologies, and take test-rides in autonomous vehicles.

The question of legal responsibility in virtual worlds also interests me. To gain a better understanding of the sort of issues that might occur, I created an avatar and participated in various virtual worlds myself. I made a female avatar, and so my experience was completely different to the real-world. This was both fun and anxiety-inducing.

From a legal perspective, if my female avatar was attacked in a virtual world, would my real-world self have incurred harm? It is an extremely difficult question.

At present, no laws exist that have been specifically designed for virtual worlds, and it is unclear whether existing laws would apply.

In fact, since there are no rules governing what country a virtual world belongs to, and since virtual worlds attract users from all over the world, it is by no means certain that Japanese laws could be applied should an incident occur. Ensuring consistency of laws internationally is therefore something I wish to explore going forward.

Encouraging technological progress while safeguarding human rights

A question that has long concerned humankind is whether laws should serve to check or to support technological progress. In the present day and age, AI is a prime example.

If the burden of responsibility on people involved in AI is too heavy, then technological development will stall. But if the burden of responsibility is too light, then human rights cannot be satisfactorily guaranteed. It is therefore vital that an optimal balance is struck between encouraging innovation and safeguarding rights.

My research stems from a desire to understand how humans are regarded by law. Civil law is largely concerned with assets and families, and so it rarely scrutinizes individuals. Since we live in an era that prizes diversity, I ask whether laws can be designed to take into account the multifaceted natures of individuals.

To take an extreme example, I explore whether laws can be extended to avatars and other persons that are at once “oneself” and “not oneself.” In this way, my research considers what laws might be appropriate when such advanced technologies become mainstream.

The book I recommend

“Roma jin no monogatari 1 Roma ha ichinichi ni shite narazu (Jo)” (Stories of the Romans 1: Rome wasn’t built in a day (Part one))
by Nanami Shiono, Shincho Bunko

Comprising 43 volumes, this epic work recounts the history of Rome in its entirety. It provides the enjoyment of a novel, and it is of great interest for me personally since it relates the lives of the people behind Roman law—which forms the basis for Japan’s civil law today. I enjoyed volumes 8–13 the most. They tell of the lives of the Caesars.

Tomoyuki Shiraishi Professor Department of Law Faculty of Law: Professor Tomoyuki Shiraishi graduated from the Faculty of Law, Keio University, and received his Ph.D. in law from the university’s Graduate School of Law. He worked as an associate professor at the Faculty of Humanities, Law and Economics, Mie University as an associate professor at the Institute of Business Science, University of Tsukuba and as a professor at the Graduate School of Social Sciences, Chiba University. Shiraishi was appointed to his current position at Sophia University in 2025.; Department of Law