In his 2019 book Online Courts and the Future of Justice, Richard Susskind explored the potential for AI to revolutionize dispute resolution. He posed an interesting question: Do litigants always want judicial decisions, or do they simply want their problems resolved swiftly, efficiently, and fairly? Susskind suggested that automated systems could, in the future, in many cases, provide quicker and more consistent outcomes than traditional courts, even if they are not perfect.

Enter Arbitrus.ai, co-founded by Brian Potts. Arbitrus.ai is providing a new platform that may bring Susskind’s vision to life—at least in the realm of arbitration. (Some of you may remember that Potts founded the company that made the keyboard for lawyers, LegalBoard, several years ago. I still have one).
Arbitrus.ai uses AI to decide contract disputes, offering a fast, cost-effective, and potentially less biased alternative to human arbitrators. The parties contractually agree to use the Arbitrus.ai platform to resolve any dispute arising out of the contract. Could this be the disruptive force that changes how businesses resolve conflicts at least of a certain level?
AI as Arbitrator: A Solution to Traditional Arbitration’s Flaws?
Traditional arbitration was meant to be a faster, cheaper alternative to litigation. But in practice, it often falls short. Potts, a practicing lawyer who became frustrated with the inefficiencies and biases in traditional arbitration, believes AI can do better.
Arbitration has become slow, often expensive, and cumbersome
Potts believes arbitration has become slow, often expensive, and cumbersome. And because arbitrators want to be hired again, they’re tempted (consciously or unconsciously) in cases involving businesses to “split the baby” to keep both parties if not happy at least not furious. When a dispute is between a business and an individual, they may lean toward the business. AI solves these problems; it’s faster, cheaper, and, if built correctly, less prone to bias.
Arbitrus.ai is designed to decide matters within the confines of a contract. Unlike litigation, where disputes can be unpredictable and wide-ranging, contractual arbitration is inherently limited to predefined issues. Potts argues that this predictability makes arbitration the perfect testing ground for AI decision-making. The contract defines the parameters of the dispute. The parties to a contract can anticipate disputes and set the framework for resolution by AI. It’s all done by agreement.
The fact it’s all done by agreement is crucial because, as Potts points out, parties can contract for any dispute resolution method they want, including AI arbitration. This contractual flexibility makes the use of AI more enforceable and practical.
Why Now? The Changing Perception of AI in Dispute Resolution
Despite the hype around AI, the idea of machines deciding legal disputes makes many uncomfortable. Skeptics argue that AI cannot grapple with complex legal, equitable, and constitutional concepts. Potts acknowledges this skepticism but believes AI is well-suited for limited and defined disputes—exactly the type that typically grows out of contracts.

This isn’t the first time AI has been proposed for dispute resolution. For years, eBay has used an online dispute resolution platform to handle millions of buyer-seller conflicts. However, Arbitrus.ai takes it further by using AI to decide more complicated matters based on contractual terms.
Challenges
Arbitrus.ai is initially targeting business-to-business disputes, where getting to a resolution quickly and cost-effectively is often more important than the specific outcome. Potts envisions a future where the platform is widely used for contract disputes, especially in scenarios where speed and cost are priorities over exhaustive legal analysis and predictable, rule-based decisions are needed. And for matters of limited exposure and liability.
The question is whether, given the speed and cost advantages of AI, is it a better way to resolve arbitrable disputes
In addition, since arbitration is designed to be private and confidential for the most part, issues about transparency that are important in judicial proceedings are less important here.
Yes, there are concerns of bias. But, as pointed out above, there are also bias issues with human arbitrators. The question is whether, given the speed and cost advantages of AI, is it a better way to resolve arbitrable disputes.
Potts is pragmatic about the limitations. He emphasizes that the goal isn’t to replace human judges or arbitrators in all contexts but to provide a faster, more efficient solution where it makes sense: “The point isn’t that AI is perfect, but on average, which is really what matters to businesses and people going through this, our system is already going to be way fairer and way more likely to get the right outcome.”
However, using the tool with consumer contracts could create issues. Many consumer agreements already contain non-negotiable arbitration clauses that may be considered unconscionable. But businesses may want to add these to these clauses a requirement that AI decides any disputes. While that may look unfair at first blush, in some circumstances, using AI to decide consumer disputes may be welcomed by consumers since it would reduce the advantage well-heeled companies have over consumers.
Potts acknowledges the potential for backlash, especially if businesses use AI arbitration to their advantage. However, he believes that transparent design and unbiased algorithms can build trust.
Other Use Cases
The platform’s potential isn’t limited to business-to-business disputes. Potts sees applications in landlord-tenant disputes, for example. These disputes clog court systems and could benefit from faster resolutions: “Imagine if we could get all the landlord-tenant disputes out of court systems.”
Potts also believes the platform could be used to advise businesses about potential outcomes before arbitration commences. The tool could be used to test out the impact and propensities for disputes of particular contractual language, to vet existing potential disputes”.
Says Potts, we could run “as many scenarios as they want about ways someone might try to breach that and run them all through our system and give you the strength of that provision and likelihood of problems. This analysis would aid in drafting contracts and reduce the likelihood of disputes.

Implications for Legal
Interestingly, Potts notes: “I talk to lawyers, and they say we would never recommend [the use of Aritrust.ai] to a client. But then I talk to GCs (General Counsels), and they’re like, Where do I sign?’”
This divergence highlights a continuing and growing tension in legal: outside lawyers are hesitant to recommend disruptive technology like AI, Gen AI, or even automation that would reduce costs and be more efficient. On the other hand, clients are interested in innovation that helps the business by reducing costs or, in the case of arbitration, getting results sooner.
Law firms naively believe that clients won’t increasingly demand the use of technology to improve efficiency and results. Like many other products, the Arbitrus.ai platform isn’t just about speeding up dispute resolution; it’s about redefining what clients expect from legal services.
The Road Ahead: Will AI Arbitration Gain Traction?
The concept of Arbitrus.ai is sound and appealing. But, like many things, the proof is in the pudding. For Arbitrus.ai to succeed, it will need to prove that it can deliver fair and accurate results that will be acceptable to users. Despite the challenges, Potts is optimistic: “We’re trying to incentivize people to agree in their contract to use this because we think that’s the most enforceable way to do this.”
Arbitration AI and Beyond
As Richard Susskind predicted back in 2019, the future of dispute resolution may increasingly reduce the level of human involvement but AI systems that provide fast, consistent, and fair decisions. By reducing the transactional costs of dispute resolution, AI could level the inherent unfairness where one side has ample resources and the other does not.
Arbitrus.ai is attempting to realize this vision, starting with arbitration. If successful, it could begin redefining the dispute resolution landscape, forcing lawyers, businesses, and even courts to rethink how justice is delivered.