Legal Minds Vs. Neural Networks: Law Schools Need to Confront Generative AI’s Increasing Sophistication in Legal Reasoning
Armin Alimardani, Melissa Porter, Warwick Gullett
Western Sydney University University of Wollongong
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
This article reports on a follow-up investigation to a 2023 study examining the academic performance of Generative AI (GenAI) in university-level Criminal Law assessment. We evaluated GenAI models from five artificial intelligence (AI) providers across two compulsory subjects in the Bachelor of Laws (LLB) degree at the University of Wollongong: Criminal Law and Procedure A (‘Criminal Law’) and Law of Torts (‘Tort Law’). For each subject, nine exam responses were generated by GenAI (n=18). Twelve of the 18 GenAI responses were blind marked by tutors who were not aware of AI involvement. The remainder were marked by the subject coordinators (authors of this study). The prompts used to generate the AI answers did not include subject-specific legal materials. GenAI achieved mean scores of 76.3% in Criminal Law and 66.0% in Tort Law, outperforming on average 82.5% and 61.0% of students, respectively. Notably, seven of the 18 AI-generated papers ranked at or above the 90th percentile of students’ performance. Compared with the 2023 baseline (mean 52.5%; percentile 22.1%), these results reflect substantial improvement in GenAI’s academic performance within two years. This study proposes a mixture of three assessment formats to ensure students demonstrate fundamental legal knowledge and skills and develop the capacity to collaborate with GenAI: (i) embedding GenAI as a legitimate component of assessment tasks; (ii) shifting written assignments, including mid-semester tasks, into AI-free invigilated formats; and (iii) a ‘relay’ approach where students either write a first draft and then use AI to analyse and suggest improvements or students prompt AI to write the first draft which students then review to verify claims and improve the output.
逐年被引趋势
暂无年度引用数据
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
社会科学Legal Education and Practice Innovations
Artificial Intelligence in Law · Artificial Intelligence in Education
参考文献 34
此处列出前 3 条