Integrity dilemmas and countermeasures of applying artificial intelligence to assist scientific research
应用人工智能辅助科研的诚信困境与应对

摘要

人工智能应用于科研辅助工作已经成为较普遍现象，同时也引发全新的学术诚信问题。应对AI辅助科研的诚信困境，需要解决学术主体与责任、科研成果可靠性与原创性、群体差异与学术公平失衡、传统治理适配不足等方面的问题。根据国内外现状与研究进展，从制度监管、技术迭代、生态影响与规范认知等角度探究原因，并提出了相应的应对路径，分别从制度、技术、教育和人文层面有效监管缓解压力，为维护健康学术生态、平衡AI创新赋能与学术诚信底线提供理论与实践支撑。

Abstract

Becoming a common phenomenon, the application of artificial intelligence in auxiliary scientific research work, at the same time, has triggered new academic integrity issues. To address the integrity dilemmas of AI-assisted scientific research, it is necessary to solve some problems such as academic subject and responsibility, reliability and originality of scientific research results, group differences and imbalance of academic fairness, and insufficient adaptation of traditional governance. Based on the current situation and research progress at home and abroad, this paper explores the causes from the perspectives of institutional supervision, technological iteration, ecological impact and normative cognition, and puts forward corresponding countermeasures. It could effectively regulate and alleviate pressures from the institutional, technological, educational and humanistic levels, providing theoretical and practical support for maintaining a healthy academic ecology and balancing AI innovation empowerment with the bottom line of academic integrity.