人智共生情境下数字媒体多模态虚假信息的算法治理

• 作者： 苏志远 ¹
• 作者单位：
  1.  中国政法大学刑事司法学院
• 发布时间：2025-01-06

摘要: 摘 要：[目的/意义]人智共生趋向下的算法供给模式，在重塑数字媒体传播生态的同时，加剧了多模态信息虚假问题的严重程度。[方法/过程]在“感官象限”的思路加持下，全面分析多模态信息的风险阶数，据此探索算法治理方案。[结果/结论]面对算法在数字媒体模态数据应用的整体风险，应当秉持源头治理的理念导向，以技术支持者和服务提供者为治理主体，分配认知安全责任。在具体措施上，通过构建系统化的算法分层监管框架，加强模块化的算法技术开发控制，细化协同化的算法服务测评方式，以期有效降低风险。

Abstract: Abstract:[Purpose/Significance] The algorithmic supply model under the trend of human-intelligence symbiosis has reshaped the digital media communication ecology, while exacerbating the severity of multimodal information falsification.[Method/Process] With the support of the "sensory quadrant" approach, we comprehensively analyze the risk order of multimodal information and explore algorithmic governance solutions accordingly. Results/Conclusion] Facing the overall risk of algorithms in digital media modal data applications, we should uphold the concept of source governance, with technical supporters and service providers as the governance bodies to allocate cognitive security responsibilities. In terms of specific measures, we aim to effectively reduce risks by constructing a systematic hierarchical regulatory framework for algorithms, strengthening modular algorithm technology development control, and refining collaborative algorithm service evaluation methods.

• 来自： 苏志远
• 分类： 信息资源管理
• 稿件状态： 已投稿期刊
• 引用： PSSXiv:202501.00935 (或此版本 PSSXiv:202501.00935V1)
  DOI:10.12451/202501.00935
  CSTR:32012.36.PSSXiv.202501.00935
• 推荐引用方式： 苏志远.人智共生情境下数字媒体多模态虚假信息的算法治理.哲学社会科学预印本平台:https://zsyyb.cn/abs/202501.00935.[PSSXiv:202501.00935V1] (点此复制)