A coming-of-age judge?

Some commentators have suggested that a solution to the judicial backlog is to increasingly rely on Artificial Intelligence (AI). The prospect of AI judges or AI-assisted judicial decision-making challenges traditional legal theories that have long been centred around human judgment, discretion, and the application of legal principles.

Countries such as Germany are already employing AI to assist in managing the backlog of cases, particularly through tasks like categorising legal documents and drafting judgments for routine cases. California has also initiated discussions on the potential use of generative AI in its courts. This, however, brings to the forefront the critical question. Can AI judge?

The integration of AI into judicial processes directly intersects with human leukocyte antigen (HLA), Hart’s legal positivism, which emphasises the law as a system of objective rules applied consistently. AI enhances legal consistency by minimising human error and subjective biases. This aligns with Hart’s vision of the law as an impersonal mechanism.

However, this potential is contingent on the quality of the training data, as AI systems are inherently dependent on historical data inputs. If these inputs reflect systemic biases—whether racial, gender, or socioeconomic—the AI’s outputs may perpetuate or even exacerbate such biases. This might undermine the rule of law’s foundational principle of impartiality. Moreover, the mechanisation of legal decision-making through AI could erode the essential judicial discretion that human judges exercise, which is necessary for contextual and moral reasoning within the legal framework.

Ronald Dworkin’s critique of legal positivism, which highlights the importance of principles and moral reasoning in judicial decision-making, raises concerns about the application of AI. While AI’s rule-based processing aligns with the positivist notion of law as a set of determinate rules, it may prove inadequate in cases requiring interpretative judgments, such as those involving constitutional morality or social justice. For instance, in the landmark case of Navtej Singh Johar v. Union of India, where the Supreme Court decriminalised Section 377 of the Indian Penal Code, the decision was not based solely on legal precedents but was deeply rooted in principles of dignity, equality, and human rights, reflecting Dworkin’s emphasis on moral reasoning. AI, constrained by its reliance on historical data and algorithmic processes, might fail to capture the nuanced ethical and moral dimensions crucial to such decisions, potentially leading to outcomes that are legally sound but morally deficient.

Another concern is rooted in the concept of legal formalism versus legal realism. Legal formalists might argue that AI, with its ability to process and apply legal rules with consistency and without emotional interference, could embody the ideal of a purely objective legal system. However, legal realists, who emphasise the importance of social context, the judge’s personal experiences, and the broader implications of legal decisions, would likely critique AI for its lack of empathy, understanding of context, and inability to consider the broader societal impact of judicial decisions. This tension reflects a fundamental challenge in integrating AI into a judicial system.

Current AI capabilities in the judiciary are largely limited to supporting roles rather than substitutive ones. AI can assist in legal research, document analysis, and even predict case outcomes based on historical data, which aligns with the concept of legal instrumentalism—viewing law as a tool to achieve social ends. However, the instrumental use of AI raises concerns about the potential reduction of law to a mere set of algorithms, stripping away the interpretative and discretionary elements that are central to many legal theories.

For instance, in the common law tradition, judges are not just interpreters of law but also creators of it through the doctrine of precedent. AI’s role in this creative process is highly questionable, as it lacks the ability to engage in the kind of nuanced reasoning and adaptation that has historically driven the evolution of legal systems.

International examples provide further insight into how AI is being integrated into judicial processes and the varying degrees of acceptance and success. In Estonia, AI is being tested to handle minor claims, reflecting a cautious approach where AI’s role is confined to low-stake decisions, which can be appealed to human judges. China’s courts, on the other hand, have implemented AI more broadly, using it to assist in routine cases such as traffic violations and small-scale theft, where legal outcomes are more predictable. These examples show that while AI can play a role in enhancing judicial efficiency, its application is still limited to areas where legal discretion is minimal and where the law is clear and unambiguous.

The potential benefits of AI in the judiciary, such as increased efficiency, consistency in decision-making, and improved access to justice, are significant. However, these benefits must be weighed against the risks of dehumanising the judicial process and the potential erosion of the rule of law. Legal theorists have long argued that justice is not merely about applying rules but about ensuring that those rules are applied in a way that is fair, just, and responsive to the needs of society. The introduction of AI into judicial processes challenges this notion by potentially reducing the role of human judgment and discretion, which have traditionally been seen as essential to achieving true justice.

In conclusion, the integration of AI into judicial processes presents both opportunities and challenges from a legal-theoretical perspective. While AI can enhance efficiency and consistency, it also raises fundamental questions about the nature of justice, the role of discretion, and the potential for bias. As legal systems around the world begin to experiment with AI, it will be crucial to ensure that these technologies are developed and implemented in a way that aligns with the core principles of justice and the rule of law. This may require a rethinking of traditional legal theories and the development of new frameworks that can accommodate the unique capabilities and limitations of AI.

The writer is Officer on Special Duty, Research, Economic Advisory Council to the Prime Minister.
Views expressed are personal.