1

Artificial intelligence (AI) is emerging as an inevitable force of change in the courtroom, reshaping how justice is administered in the digital age. With the launch of the Judicial AI Committee, South Korea’s judiciary has begun laying the groundwork for the formal adoption of AI tools to enhance efficiency and public access to justice. The initiative aims to tackle chronic trial delays while reducing administrative burdens on judges.

The judiciary’s focus on AI stems from a clear motivation — resolving the longstanding problem of drawn-out trials. Korean courts face an ever-growing caseload, with judges and clerks overwhelmed by procedural steps, documentation, and legal research. These repetitive, time-consuming tasks have prolonged proceedings, feeding public distrust in the system and the notion that “justice delayed is justice denied.”

AI is now viewed as a powerful solution. By scanning vast databases of precedents and legal documents, AI can identify similar cases, summarize core issues, and automatically organize key legal arguments for judges. It can also extract crucial sentences from lengthy case records or generate draft rulings modeled after prior judgments — thereby easing the burden of decision writing. AI-assisted anonymization of personal data is also being introduced to speed up publication of rulings.

Attorney Kang Min-gu, a former judge of 36 years and now with the law firm Doull, said, “Introducing AI into the judiciary will transform not just efficiency, but the very culture of litigation itself. The biggest impact will come from automating judgment writing, which will help eliminate trial delays.”

“The main cause of delay today is judges struggling under the workload of writing rulings,” Kang explained. “Even writing three judgments a week causes backlogs, delaying verdicts for months. If AI handles record summaries, argument organization, and evidence analysis, judges can focus on essential reasoning. When AI automatically provides summaries of claims and relevant precedents, judges could draft rulings as if assembling Lego blocks. That means handling five or six cases a week instead of three — with fewer errors and much faster sentencing.”

AI is also expected to expand access to justice. Citizens could use AI tools to quickly locate relevant case law, legal principles, and document templates. The courts plan to integrate voice recognition, auto-captioning, and image-based communication aids to help people with disabilities and improve inclusivity.

Choi Kyung-jin, president of the Korean Society for AI Law and professor at Gachon University, said, “The greatest promise of AI in law is strengthening citizens’ right to a fair trial. It allows for more consistent sentencing in similar cases, bringing about a fairer and more equal justice system.”

However, challenges remain — particularly around legal accountability and ethics. When AI-generated analyses inform judicial reasoning, the question arises: who bears responsibility for an error? In law, where a wrong decision can have grave consequences for litigants, ambiguity over accountability poses serious risks. Bias in training data is another concern; an AI trained on regionally or historically skewed cases might replicate social prejudices, undermining fairness and neutrality.

Procedural fairness also comes into question. If AI influences sentencing or evidentiary assessments, human judges may lose time for deliberation, and algorithmic reasoning could eclipse human judgment. Critics warn that overreliance on technology might usher in an era of “mechanical justice,” stripped of empathy and moral insight.

Still, Professor Choi remains optimistic: “Legal and ethical dilemmas exist for humans too. What matters is continuously reducing those risks. With a human-in-the-loop (HITL) system — where humans oversee AI training, adjustment, and verification — such concerns can be effectively managed.”

Globally, judicial systems are moving in the same direction. U.S. state courts are experimenting with AI chatbots, recidivism prediction tools, and automated ruling systems. Singapore has introduced generative AI on a pilot basis in its small claims courts, while Australia launched AMICA in 2020 — an AI service helping couples resolve family and custody disputes.

In April, South Korea’s Supreme Court established the Judicial AI Committee as an advisory body under the Court Administration Office, effectively creating a control tower for AI integration in the judiciary. The committee will define the direction of AI use in the courts, oversee related development projects, and review legal frameworks. By the end of this year, it plans to unveil a national AI roadmap for the judiciary and a series of mid-to-long-term pilot projects.