International Open Access Journal Platform

生成式人工智能的个人信息安全风险和刑法完善路径

Personal Information Security Risks of Generative Artificial Intelligence and the Path to Criminal Law Improvement

Progress in Social Sciences / 2025,7(7): 547-553 / 2025-07-21 16 10

View Download PDF

• Authors: 邵颖 瞿丹文
• Information:

  上海政法学院，上海

• Keywords:
  Generative artificial intelligence; Crime of infringing upon citizens' personal information; Video generative artificial intelligence
  生成式人工智能; 侵犯公民个人信息罪; 文生视频
• Abstract: With the further advancement of generative artificial intelligence technology, various generated products such as text, images, and videos pose diverse risks and challenges to citizens' personal information security. Currently, at the criminal law level, illegal acquisition and disclosure of citizens' personal information by generative artificial intelligence can only be punished as the crime of infringing upon citizens' personal information. However, there are still many loopholes in the criminal law's regulation of such crimes, including chaotic criminal liability systems, excessive use of publicly available personal information beyond reasonable scope, and insufficient protection of biometric information. These issues urgently require criminal law to respond one by one. In order to better apply generative artificial intelligence to human life and safeguard citizens' information security rights, criminal law must promptly make adjustments at both the legislative and judicial levels, timely modify the types of criminal acts, adjust the liability principles for generative artificial intelligence crimes, strengthen the protection of publicly available personal information and biometric information, and build an integrated governance model of administrative and criminal law to promote the unlimited development of the future artificial intelligence industry. 随着生成式人工智能技术的进一步提升，文本、图片、视频等不同生成物给公民的个人信息安全带来多样的风险挑战。目前在刑法层面对于生成式人工智能非法获取、泄露公民个人信息的行为只能以侵犯公民个人信息罪定罪处罚，但是刑法对该类犯罪的规制仍然存在诸多漏洞，刑事归责体系混乱、已公开个人信息超过合理使用范围、生物识别信息保护不足等问题亟需刑法一一作出回应。为了使生成式人工智能更好地应用于人类生活和维护公民信息安全权益，刑法必须在立法和司法层面及时作出调整，适时修改犯罪行为类型，调整生成式人工智能犯罪的归责原则，加强对已公开个人信息和生物识别信息的保护，构建行刑一体化的治理模式，以促进未来人工智能产业的健康发展。
• DOI: https://doi.org/10.35534/pss.0707093
• Cite: 邵颖，瞿丹文．生成式人工智能的个人信息安全风险和刑法完善路径［J］．社会科学进展，2025，7（7）：547-553.