GenAI模型赋能社会救助敏捷治理的伦理风险与实践进路OA

The advancement of Generative AI models has opened new pathways for agile governance in social assistance.This study elucidates the technical advantages of GenAI in enhancing the precision,responsiveness,and cross-departmental collaboration of social assistance,as well as its intrinsic alignment with the concept of agile governance.Furthermore,it critically examines three ethical challenges arising from the integration of GenAI into social assistance systems,including blurred accountability boundaries leading to an accountability vacuum,overreach of technical actors resulting in the alienation of rule-making,and erosion of humanistic care by instrumental rationality.To address these challenges,this study draws on agile governance theory and the ethics of responsibility to reveal the"efficiency-accountability"tension embedded in GenAI-driven social assistance processes.It proposes a three-dimensional practical approach centered on the re-anchoring of administrative responsibility,the balancing of technical power,and the reintegration of humanistic rationality.By constructing a governance paradigm characteristic of synergistic evolution of"institutions,technology,and values,"this study aims to provide theoretical support for the digital transformation of social assistance that balances technical feasibility with governance robustness.