<![CDATA[Indonesia’s National Criminal Code (Kitab Undang-Undang Hukum Pidana, KUHP), enacted through Law No. 1 of 2023, became effective on 2 January 2026, creating a new learning environment in which students must master recently operative statutory rules while generative artificial intelligence (GenAI) is already embedded in everyday academic practice. This study examines the perceived pedagogical utility of GenAI for learning the KUHP among 82 university students studying in Jakarta who were surveyed during January-February 2026. A cross-sectional mixed-methods design combined ten five-point Likert items with three open-ended questions. Seven items measured AI-supported learning utility, two measured source-verification literacy, and one measured overall perceived effectiveness. The seven-item learning-utility scale showed high internal consistency (Cronbach’s alpha = .943) and a mean of 3.68 (SD = .76, 95% CI [3.51, 3.85]). Students rated AI most positively for explaining previously difficult KUHP provisions (M = 3.84), while ratings were lower for identifying the legally relevant provision in a case (M = 3.56) and strengthening legal argumentation (M = 3.57). Learning utility correlated strongly with overall perceived effectiveness (Spearman rho = .721, p < .001), as did verification literacy (rho = .755, p < .001). In an exploratory ordered-logit model, both standardized learning utility (OR = 6.54, p < .001) and verification literacy (OR = 8.84, p < .001) independently predicted higher overall effectiveness ratings after adjustment for semester group. Open-ended responses emphasized simplified understanding, case application, and faster information access, but also indicated the need for official-source verification. The findings support a bounded pedagogical role for GenAI: it is most defensible as a scaffold for explanation, examples, and formative feedback, not as a substitute for statutory reading or independent legal judgment.]]>
Copyrights © 2026
