首页|人工智能"逐案设法"治理模式的优化

人工智能"逐案设法"治理模式的优化

扫码查看
面对人工智能技术迭代浪潮,我国倾向于针对特定技术出台"定制型"规范.这种"逐案设法"以多个部门联合制定的部门规章作为治理载体,能够迅速使新兴领域实现"有法可依",但存在着规制范围重合、规制内容冲突及立法资源浪费等问题."逐案设法"反映出行政主管部门的技术风险焦虑,是"万事皆有法式"传统理念的再现.为此,需要摒弃"零风险"的规制目标与"法律万能主义"的思维程式,优先借助行政指导、企业承诺及行业标准等柔性治理方式.立法过程应着重解决事物新特性与传统监管手段"不匹配"问题,通过引入立法评估促进规范的动态调整,以应对技术的快速发展与更新迭代.
The"Case-by-Case Lawmaking"Governance of Artificial Intelligence:Problems and Optimization
In response to the rapid iteration of Al technologies,China has tended to adopt a"small incision"legislative approach for specific technologies and their application areas,resulting in"customized"norms.This"case-by-case legislation"are characterized by departmental regula-tions jointly formulated by multiple departments,ensuring that emerging fields have legal backing.However,"case-by-case legislation"lacks the necessary justification,and is likely to lead to overlap in the scope of regulation,redundant legal concepts,conflicting regulatory content,and wasteful leg-islative resources.The case-by-case legislation reflects the technological risk anxieties of administra-tive departments,a reiteration of the traditional"law for every scenario"ideology.To address this,it is imperative to abandon the"zero-risk"regulatory goal and the"omnipotence of law"mindset,making use of"soft law"such as administrative guidance,corporate commitments,industry stand-ards,etc.The"case-by-case approach"should focus on exploring and solving the problem of"mis-match"between the new characteristics of the matter and the traditional means of regulation.Legis-lative evaluation should be introduced to encourage dynamic adjustments and optimization of norms,to cope with the rapid development and iteration of AI technologies.

AIdepartmental regulations,specialized legislationexperimental legislationle-gal regulationsoft law

邓建鹏、马文洁

展开 >

中央财经大学法学院 北京100081

人工智能 部门规章 专门立法 试验性立法 软法

中央高校基本科研业务费项目(中央财经大学重大研究支持计划)

024151623003

2024

南京社会科学
南京市社会科学界联合会 南京市社会科学院 中共南京市委党校

南京社会科学

CSTPCDCSSCICHSSCD北大核心
影响因子:0.998
ISSN:1001-8263
年,卷(期):2024.(6)
  • 24