Express Justice: Would You Trust Your Case to an AI?

View on LinkedIn

Imagine someone owes you $5,000. A significant sum, but not large enough to justify three years of litigation. The legal fees, the time invested, and the emotional toll would likely exceed the amount of the debt. As the saying goes, the cure would be worse than the disease.

This reality presents a troubling dilemma:

What would you do if you were offered an alternative? An AI that promises to resolve your case in 20 days, with 90% accuracy. The catch: the decision is final and unappealable. No second chances. No possibility of introducing new evidence. No judge to reconsider the arguments.

Last week I conducted a LinkedIn poll on the use of artificial intelligence in justice institutions. The results were revealing: 50% of participants see AI as a tool for streamlining processes, while only 11% expressed concerns about its safety. A significant 39% believe its usefulness depends on the context.

https://www.linkedin.com/feed/update/urn:li:activity:7286527334621331458/

These numbers suggest something deeper than mere digital impatience. When the traditional judicial system becomes so slow that it effectively denies access to justice, are we not already living within a failed system? The difference is that instead of a definitive "no" in 20 days, we receive a "maybe" that drags on for years.

In the United Kingdom, the tool Söze demonstrates the potential of AI: it can analyze in 30 hours what would take a team of researchers 81 years. Colombia, for its part, has already established a regulatory framework that integrates AI as a support tool, while preserving human judgment in final decisions.

But these advances, though promising, are merely palliatives for a system that needs deep structural reform. The real problem is not only the slowness, but how that slowness systematically favors those with the resources to sustain lengthy judicial proceedings.

When the cost and duration of a judicial proceeding exceed the value in dispute, the justice system becomes a luxury, not a right. This reality drives many to consider alternatives that in other circumstances would seem unthinkable — such as entrusting an irrevocable decision to an algorithm.

The COMPAS system in the United States warns us about the dangers of algorithmic bias. But is our current system, with its economic and temporal barriers, not also biased against those who cannot afford to wait years for a ruling?

What does it say about our judicial system that the idea of an unappealable algorithmic decision with a 10% margin of error might seem preferable to a traditional legal process?

Perhaps the true disruption we need is not technological, but structural: a judicial system that does not force us to choose between fast justice and fair justice.

In the meantime, let us return to those $5,000. Perhaps the answer does not lie in choosing between swift justice without guarantees and guaranteed justice that is out of reach. Perhaps the first step is to explore these automated systems in small-claims cases, where the cost of the traditional process makes the pursuit of justice unviable.

Minor claims could be the perfect laboratory for these innovations: the risk is limited, but the potential to improve access to justice is enormous. If we can make it work here, we will have taken the first step toward a judicial system that combines the best of technological efficiency with the procedural guarantees that a democratic society demands.