AIGC检测介入学位论文AI代写的技术审视与法理反思OACHSSCD
Technical Scruting and Jurisprudential Reflections on AIGC Detection in Preventing AI-Generated Thesis Misconduct
《中华人民共和国学位法》规定学位论文代写行为构成学术不端,应不授予学位或撤销学位.高校普遍将AI代写视为学术不端,并以第三方机构的AIGC检测结论作为衡量基准.从技术原理看,现有AIGC检测工具主要基于水印技术、训练分类器、零样本检测等,受困于检测技术限制与AI幻觉,当前检测能力难以真正反映真实AIGC疑似水平.从法理角度看,学位论文中对AI的利用需区分工具性使用与创造性使用,AI代写构成学术不端需满足比例性原则约束,但AIGC检测工具无法厘清这些边界;由于检测结论真实性存疑、不符合法定证据形式,因而证据力不足;AIGC检测结论的证明力薄弱,单独使用难以满足行政执法与行政诉讼的证明标准.为此,立法机关、教育行政管理部门需要构建对第三方商业性AIGC检测平台的规制体系,综合考量其检验原理、检验能力和实施准入机制,引导高校理性选择AIGC检测工具.高校应审慎对待第三方机构的AIGC检测结论,明确其仅可用于辅助材料使用,且要有充足证据验证结论的真实性;落实以人为本的人工核验制,由学位委员会判断学位论文是否构成"AI代写型"学术不端;构建AIGC学术性使用诚信机制,实施被动检验与引导学生主动披露并行,教育学生正确、合理使用人工智能工具.
Law of the People's Republic of China on Academic Degrees stipulates that ghostwriting in degree theses constitutes academic misconduct,warranting the withholding or revocation of degrees.Universities generally regard AI ghostwriting as academic misconduct and rely on AIGC detection results from third-party institutions as the primary benchmark.Technically,the existing AIGC detection tools primarily utilize watermarking techniques,trained classifiers,and zero-shot detection.However,hampered by technical limitations and"AI hallucinations,"their current detection capabilities struggle to accurately reflect the actual likelihood of AIGC.From a jurisprudential perspective,the utilization of AI in degree theses requires distinguishing between"instrumental use"and"creative/substantive use."Classifying AI ghostwriting as academic misconduct must satisfy the constraints of the proportionality principle,yet AIGC detection tools cannot clarify these boundaries.Furthermore,due to questionable authenticity and non-compliance with statutory forms of evidence,the probative value of AIGC detection reports is insufficient.Their weak evidentiary strength also makes them inadequate,when used alone,to meet the standard of proof required in administrative enforcement and administrative litigation.In response,legislative bodies and educational administrative departments need to establish a regulatory framework for commercial third-party AIGC detection platforms.This includes implementing an access mechanism based on comprehensive evaluation of their detection principles and capabilities,guiding universities in the rational selection of AIGC tools,and preventing market disorder.Universities should treat third-party AIGC detection conclusions with caution,explicitly defining them for auxiliary use only and requiring sufficient evidence to verify their authenticity.A human-centric,manual verification system must be implemented,empowering academic degree committees to determine whether a thesis constitutes AI ghostwriting as academic misconduct.Additionally,it is crucial to build an integrity mechanism for the academic use of AIGC,implementing a dual strategy of passive inspection and actively guiding students to disclose AI use,thereby educating them on the correct and appropriate utilization of artificial intelligence tools.
王家睿
山东大学 法学院,青岛 266237
社会科学
AIGC《中华人民共和国学位法》学术不端代写行为
AIGCLaw of the People's Republic of China on Academic Degreesacademic misconductghostwriting
《重庆高教研究》 2026 (4)
71-81,11
司法部法治建设与法学理论研究重点项目"习近平法治思想引领司法学自主知识体系建构研究"(25SFB1001)
评论