By Philip Winkle

Introduction

If you grew up watching science fiction, you probably expected the robot takeover to resemble The Terminator, I, Robot, or The Matrix. But, instead of battling automatons or desperately hacking the algorithm trying to override humanity’s destiny, this Brave New World is really just you—someone working in or studying international arbitration—staring at an arbitration software using Artificial Intelligence (AI) that can do your job cheaper and faster.

The future is less exciting, but perhaps more insidious than those cautioned by Aldous Huxley or the Wachowskis. You are still stuck in the Matrix, but sans the ability to fly. The only societal castes are the employed and unemployed, and you are left wondering if that $300,000 law degree was worth the investment.

Assessing whether AI will cause human arbitrators to go the way of the dinosaur or the floppy disk is worth examining and not merely as an exercise in quelling anxiety. The tension surrounds not just whether AI replaces humans but how humanity can adapt.

To resolve these tensions, this blog will first examine AI’s likelihood of fully replacing human arbitrators. Then, it will discuss the challenges ahead, covering how governments, law firms, institutions, and law students can adapt effectively.

Part I: Is AI Replacing Human Arbitrators?

A major appeal of arbitration, particularly international arbitration, is its utility in resolving complex disputes more efficiently and cost-effectively than litigation. Following this rationale, a cheaper and faster AI-powered program may seem like a threat. Tools like ClauseBuiler AI, ArbiLex, and CoCounsel have already shown to be promising time savers and useful tools for practitioners. However, casual observers of these tools may fundamentally misunderstand AI’s current and future use in arbitration in two important ways.

The first misunderstanding concerns the nature of AI itself. AI is essentially the science and engineering behind intelligent computer programs, and the categories of AI involved are either Narrow AI or Strong AI. Narrow AI refers to a system’s ability to perform and mimic tasks in ways similar to humans, while Strong AI encompasses the capacity for imagination and even consciousness. Applications for AI include language processing, problem-solving, and speech recognition programs. Recent advancements owe their success to the concept of Machine Learning (ML), which describes the inductive approach that uses algorithms and preset data to imitate human decision-making. A subset of ML, known as Deep Learning (DL), goes a step further, allowing the program to ingest layers of data, even raw datasets, to produce an output.

Form follows function, and AI’s structure helps predict its role in arbitration. While self-driving cars and language programs are impressive, their reliance on large datasets is noteworthy. Success in AI is thus far largely limited to Narrow AI. The need for predetermined data is a limiting factor in any industry, let alone arbitration.

The second misunderstanding relates to the value of human arbitrators, which can be often overlooked at one’s peril. Efficiency and cost are frequently touted as virtues of arbitration for businesses looking to avoid litigation. However, AI raises the question of whether humans add intrinsic value to arbitration. Informed by experience and having the capacity for creativity and emotional intelligence, humans can better adapt and plan around the unique dynamics of a dispute that are not captured in any data set that an AI relies on. Whereas humans can assess the unique personalities, relationships, and even personal agendas involved in a dispute, AI will respond to generalized information. Moreover, individuals may distrust the fairness of AI altogether. After all, AI’s reliance on data sets and algorithms can introduce issues of bias and trust that humans may be better equipped to address by responding to concerns directly.

Taken together, it is unlikely that AI will fully replace human arbitrators. AI is limited in how it operates, while human arbitrators can look beyond the data and use their soft skills to resolve disputes. Therefore, instead of replacing humans, AI is more likely to augment their abilities.

“Assessing whether AI will cause human arbitrators to go the way of the dinosaur or the floppy disk is worth examining and not merely as an exercise in quelling anxiety.”

Philip Winkle

Part II: What Next?

Humans in arbitration are here to stay, but so is AI, and future arbitrators must adapt to succeed. To achieve this, governments, law firms, institutions, and law students each have a role to play in shaping the future.

Governments must actively pursue legislation to better understand and regulate this rapidly growing industry. Universities, companies, and private-sector associations can play a unique role in researching and developing next-gen AI tools and educating the next generation of practitioners. Law firms, in turn, must not only incorporate AI into practice but also establish ethical guidelines for how AI should be used by practitioners. Finally, law students must recognize that they will be the practitioners utilizing AI throughout their careers. If the next generation ignores AI and its applications now, they swap the problem of being replaced by AI with another problem—being overlooked by employers. Rather than being replaced by AI, they could be left behind by those who understand how to integrate the use of AI into their practice effectively.

Conclusion

In conclusion, the future is potentially less apocalyptic than the Terminator, but society isn’t getting off scot-free. Creating tools, producing legislation, developing guidelines, and actively using AI in practice are the next steps toward a more comfortable, albeit duller future. The Brave New World may not be made for the movie screen, but it is made for humans—at least until the aliens invade.