Reflection and regulation of intelligent recommendation algorithms in short video addiction
Yang Meiling，Pan Jing
College of Humanities, Beijing University of Posts and Telecommunications
Abstract： The proliferation of short video addiction is closely associated with the deep-seated application of intelligent recommendation algorithms, emerging as a significant issue affecting individual development and social governance. Intelligent recommendation algorithms continuously capture users′ attention through the technical closedloop of "data profiling—enhanced push—instant feedback", thereby inducing systemic public health risks. The current governance framework shows remarkable lag at the algorithmic level, mainly manifested in vague criteria for algorithmic liability determination, insufficient legislation, and fragmented regulatory mechanisms. Therefore, the governance of intelligent recommendation algorithms in short video addiction should adopt a laddered governance path based on technical characteristics, guided by social impacts, and guaranteed by legal responses: at the legal level, clarify the standards for identifying addiction, and establish the algorithm risk classification and three-party liability system; at the regulatory level, establish a dynamic monitoring mechanism for algorithms, and implement regulatory sandboxes and stepwise interventions; at the multi-stakeholder co-governance level, improve industry self-regulation, smooth channels for public participation, and promote the alignment of international rules. It is expected to provide practical approaches for tackling algorithm-driven addiction and offer paradigmatic reference for the legal regulation of technological risks in the digital era.
Key words : short video; addiction; intelligent recommendation algorithm; legal regulation; multistakeholder governance
