 Usos, percepciones y gobernanza de los sistemas de inteligencia artificial en la Rama Judicial: hallazgos de tres encuestas en ColombiaJuan David Gutiérrez and
David Stiven Peralta
No 21349, Documentos de trabajo from Escuela de Gobierno - Universidad de los Andes Abstract: This document presents the main findings of three surveys conducted among judicial officials in Colombia regarding the use, perception, and governance of artificial intelligence (AI) tools in the justice sector. The surveys were conducted through three electronic forms between February and July 2024. The first two surveys were answered by judicial officials who took part in the course “Artificial Intelligence for the Administration of Justice: Fundamentals, Applications, and Best Practices”, offered jointly by the Universidad de los Andes and the Rodrigo Lara Bonilla School of the Superior Council of the Judiciary. The first survey was answered at the beginning of the course by 1 391 judicial servers, while the second was completed at the end of the course by 824 servers. The third survey, designed and administered directly by the Superior Council of the Judiciary, received responses from 3 152 judicial servants. We identified nine main findings from processing the survey data. First, around one third of judicial officials surveyed reported using AI tools to facilitate their work, with ChatGPT being the most popular tool (38% of users), followed by Copilot (20%) and Gemini (11%). However, the frequency of AI tool use is relatively low: at the beginning of the course, 10.5% of respondents reported using these technologies daily or weekly. The course appears to be correlated with positive effects in terms of the use of these technologies: the reported use of generative AI tools increased from 26% to 44%, and the frequency of daily or weekly 4 The authors would like to thank Assistant Judge Nathalia Gaona and researcher Sebastián Hurtado (Universidad de los Andes) for their comments and criticism of a preliminary version of this report. Any errors or omissions in the document are the sole responsibility of the authors. This research was funded by the Alberto Lleras Camargo School of Government at the Universidad de los Andes. 5Juan David Gutiérrez, PhD, is an associate professor at the Alberto Lleras Camargo School of Government at the University of the Andes. Email: juagutie@uniandes.edu.co 6David Stiven Peralta M., MSc, is a research assistant at the Alberto Lleras Camargo School of Government at the University of the Andes. Email: ds.peralta@uniandes.edu.co use of AI tools increased by 14 percentage points upon completion of the course. Secondly, the most frequent uses of these technologies by judicial staff include case law search (59%), legislation search (49-52%), meaning and definition search (46-51%), and text summarization (40-47%). Third, 79-85% of the judicial officers surveyed accessed free versions of AI tools, raising concerns about the protection of personal data and information confidentiality. Fourth, 77-89% of respondents identified potential negative aspects associated with the use of Generative AI tools for the administration of justice, the main ones being: protection of confidentiality and personal data (76-82%), problems with the quality of results (62-65%), and copyright infringement (55-59%). Additionally, the course appears to have contributed to a 12-percentage-point increase in awareness of ethical risks. Fifth, the main barriers identified by judicial officials to the use of AI tools in the administration of justice are a lack of training (81%), the absence of institutional guidelines (67%), and access difficulties (61%). Ninety-one percent had not received formal training, but 93% considered it necessary. Sixth, contrary to common perceptions, judicial officials requested regulatory frameworks: 75-88% believed that mandatory rules are necessary for the use of AI by judicial officials, and 76-87% believed the same for lawyers. Between 79% and 92% considered it useful to establish official rules on the use of AI in the justice sector. Seventh, despite the risks identifi Keywords: inteligencia artificial; algoritmos, modelos de lenguaje a gran escala; chatbots; justicia; jueces; tribunales; servidores judiciales; Colombia (search for similar items in EconPapers) Downloads: (external link) Related works: Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text Persistent link: https://EconPapers.repec.org/RePEc:col:000547:022182 Access Statistics for this paper More papers in Documentos de trabajo from Escuela de Gobierno - Universidad de los Andes Contact information at EDIRC. |
Is your work missing from RePEc? Here is how to contribute.
Questions or problems? Check the EconPapers FAQ or send mail to .
EconPapers is hosted by the
School of Business at Örebro University.