Artificial intelligence and alternative dispute resolution are both rapidly expanding in the legal field. AI aims to increase efficiency through technological advancement, while ADR achieves this through procedural innovation.
As AI capabilities improve, the convergence of AI and ADR is inevitable, and this could fundamentally change our legal system.
In the US, AI in ADR has generally been limited to document analysis, legal research, case argument evaluation, offer generation, marketing and billing. However, the full realization of AI’s potential in the legal field is the creation of an autonomous AI legal system that functions as an alternative to our traditional legal system. No human judges, juries or lawyers.
This new form of dispute resolution would consist of AI agents representing litigants and AI decision makers adjudicating cases. Disputes could be resolved in minutes, decision makers could be coded with extensive knowledge of the law and objectivity, and litigants would receive equally competent representation. Under this system, the long-sought ideals of blind justice and equal protection of the law could be more attainable than ever.
The convergence of AI and ADR reflects a natural progression driven by clients looking to reduce litigation time and costs. Although autonomous AI legal systems are still years away, the path to this endpoint could soon raise fundamental questions about the necessity and effectiveness of traditional legal institutions.
Compared to backlogged traditional courts, an autonomous AI court system could provide a highly efficient method of resolving disputes. Parties would first submit their claims, facts, defenses and desired damages to an AI-powered online platform. The claimant would be supported by an AI agent, who would review the details of the case, consult with the claimant, gather additional information, research legal precedents and present the strongest possible argument.
The suspect’s AI agent would take the same steps when preparing the defense. Once both sides have prepared their cases, the AI decision maker assesses the arguments and makes a decision – possibly within minutes or even seconds. While the idea of a legal system without human lawyers may be unsettling, this new form of ADR could level the playing field. AI decision-makers and AI-powered legal representation could create a justice system that is truly impartial.
Free from political pressure, coded for maximum objectivity and equipped with extensive legal knowledge, AI decision makers could make more fair and consistent rulings than humans. Litigants could receive equal, competent representation, thereby promoting fairness. By reducing delays, an AI justice system could achieve the speedy justice that our current justice system often struggles to achieve.
However, there are three main issues that hinder the development of an AI legal system.
First, the technology needed to create an autonomous AI legal system is not available. AI-powered tools still fall short of the capabilities of human lawyers. Large language models remain susceptible to hallucinations and biases, are unable to handle complex legal investigations from start to finish, and lack the ability to empathize or understand the human aspects of legal disputes.
Second, the process of refining the LLM model that forms the basis of the AI decision maker would lead to widespread debate and controversy. The core problem lies in the creation of the AI decision maker and the data used to train it, which would determine whether it can produce fair results.
Bias in training data can inadvertently influence decisions, and the choice to train LLMs based solely on legal precedents or to incorporate cultural, social, and local norms raises further questions about fairness and justice. Furthermore, AI legal systems may struggle to gain legitimacy and public trust. Without community support and a robust system for appeals and oversight, AI-driven decisions could be met with resistance and deemed unlawful.
And third, the potential for crimes in an AI legal system is undeniable. Bad actors with access to algorithms, data, and codebases that power the AI platform can alter the inner workings of LLM models, influence case outcomes, and sow distrust.
Assuming technological advances address these challenges, the first adopters of an AI legal system would likely be local governments, developing countries, or large corporations.
Local governments could use AI legal systems for simple, non-controversial disputes, such as parking tickets or minor civil violations. In developing countries, AI legal systems could provide more reliable, transparent justice and combat corruption.
By providing consistent and fair dispute resolution, these systems could promote stronger institutional values and encourage foreign investment. But this would come with many challenges as developing countries generally lack the resources, technological capabilities and capacity to implement such new technology.
In the government and country context, AI legal systems could be introduced in a hybrid model, where AI platforms support human decision makers. Long testing phases would refine the AI system, allowing gradual improvements and assessments before a full transition.
It is more likely that the very first users of AI dispute resolution platforms will be large corporations – and, by extension, unwitting consumers – due to the strong financial incentives. It is already common for companies to include ADR clauses in consumer contracts, forcing consumers to resolve disputes through arbitration rather than traditional courts.
Two recent high-profile examples where the Walt Disney Co. And Uber Technologies Inc.that forced injured and deceased consumers into arbitration highlights the lengths companies would go to keep lawsuits out of traditional courts.
At the same time, large companies are rapidly integrating AI in their operations to reduce costs and increase efficiency. Given these incentives and aggressive approaches to mitigating legal risks, companies could soon embrace AI-powered dispute resolution platforms as a cost-effective and risk-reducing alternative.
Is the promise of faster, cheaper access to justice and equal representation worth the trade-off of removing human decision makers from the legal process? As we approach the possibility of autonomous AI legal systems, such ethical questions are also inevitable.
This article does not necessarily reflect the views of Bloomberg Industry Group, Inc., the publisher of Bloomberg Law and Bloomberg Tax, or their owners.
Author information
Oliver Roberts is co-head of Holtzman Vogel’s AI practice group at and CEO and co-founder of Wikard, a legal AI technology company.
Write for us: Author Guidelines