类ChatGPT生成式人工智能的运行机制、法律风险与规制路径

The Operation Mechanism,Legal Risks and Regulatory Methods of Generative Artificial Intelligence like ChatGPT

生成式人工智能通过“准备-运算-生成”的运行模式为人们提供了创造性和高效的内容生成工具,深刻改变了创作、沟通和信息传播方式。但其在给人们带来便利的同时也带来了一系列的法律风险,如在准备阶段出现了数据安全和算法偏见风险等问题,在运算阶段存在隐私泄露和网络侵权、欺诈风险等问题,在生成阶段引发了版权及版权归属不明和商业秘密泄露及其导致的不正当竞争等问题。对于以上法律风险问题,需要以安全负责和科技向善为着眼点,选择相对应的法律规制路径,这有助于平衡技术创新与社会稳定之间的关系,保障公众权益。

Generative artificial intelligence provides people with creative and efficient content generation tools through the operation of the three phases of“preparation-computation-generation”,which has profoundly changed the way of creation,communication and information dissemination.However,while bringing convenience to people,it also brings a series of legal risks.The preparation stage has led to problems such as data security and the risk of algorithmic bias;the computing stage has highlighted problems such as privacy leakage and network infringement,and fraud risk;and the generation stage has triggered problems such as unclear copyright and copyright attribution,and leakage of trade secrets and the unfair competition they cause.With regard to the legal risks arising in the three major phases,it is necessary to put forward corresponding legal regulation paths with a focus on safety and responsibility and technology for the good,which will help to balance the relationship between technological innovation and social stability,and safeguard the public interest.