Colombia: The Use of ChatGPT in Drafting Judicial Rulings

Author:

Erica Giraldo Toro - Lawyer & Managing Director of iGlobal.Lawyer

Abstract

As already discussed for some time in this blog, Artificial Intelligence (AI) is increasingly taking up significant space in all areas, both personal and professional. In this brief article, we will be focsuing on a recent case that took place in Colombia, because one of the legal issues in South America now revolves around judges using ChatGPT to draft their judicial rulings.

From Chat to Verdict: ChatGPT Enters the Courtroom

In 2023, a notable case emerged in Colombia involving a judge's ruling on health services for a child with autism. The judge, Juan Manuel Padilla, issued a second-instance decision ordering the fulfillment of health services required by the autistic child enrolled in a health insurance plan. What made this case particularly noteworthy was the judge's (unconventional) approach to drafting the ruling: he incorporated responses from ChatGPT into his legal reasoning.

Regarding this case, an action was filed against the ruling, which involved the use of AI before the Constitutional Court on August 16, 2024.

The Court concluded that AI was not used improperly because, although the answers provided by ChatGPT were included in the reasoning, they were not relevant to the decision-making concerning the child’s rights.

However, this led the Constitutional Court to warn that AI cannot be used to replace the human logical reasoning that each judge must apply when interpreting facts, evaluating evidence, reasoning, and making a decision. If used improperly, it could lead to a violation of the guarantee of a natural judge and due process. The Court further added that AI could be used in administrative and document management areas and in correcting summaries and texts, provided that rights such as habeas data or privacy are not infringed.

The Court also urges all judges in the country to assess the proper and ethical use of this tool. The use of ChatGPT and other similar tools can be helpful, but their improper use can also lead to serious violations of various rights to which a human being is entitled.

The creator of this tool, Sam Altman, who spoke before the U.S. Senate, acknowledged that OpenAI systems could be destabilizing and stated: “We think that regulatory intervention by governments will be critical to mitigate the risks of increasingly powerful models.”

Not only can judicial decision-making not depend on AI due to issues of reliability and purely human reasoning, but we must also remember that justice is not just logical reasoning—emotions play an important role in decision-making. Additionally, judges must currently develop concrete skills to evaluate whether evidence is genuine; we must not forget that AI can fabricate evidence, such as cloning voices or creating plausible documents with non-existent research.

AI is making significant strides, and nations must be prepared for this new era—we are talking about a double-edged sword.

Bibliography

Constitutional Court of the Republic of Colombia, 2024. Judgment T-323/24, Presiding Magistrate Juan Carlos Cortés González.

Available at: https://www.corteconstitucional.gov.co/relatoria/2024/T-323-24.htm (Accessed: 17 November 2024).

Fung, B., 2023. "OpenAI CEO Sam Altman testifies before Congress".

Available at: https://edition.cnn.com/2023/05/16/tech/sam-altman-openai-congress/index.html (Accessed: 15 November 2024).

Sciences Po, 2023. "Sciences Po bans the use of the ChatGPT tool".

Available at: https://newsroom.sciencespo.fr/sciences-po-interdit-lutilisation-de-loutil-chatgpt/ (Accessed: 17 November 2024).