生成式AI迅速发展，其应用在收到积极推广的同时也饱受争议。目前对生成式AI接受度的研究仍然较少，本研究从影响接受度的拉力和推力以及人-机-环境系统的视角出发，探究威胁感知和能力感知对接受度的核心预测作用，以及人类身份差异、AI身份差异、任务价值和任务类型对感知和态度的影响。研究1证明了不同的威胁感知和能力感知对接受度的影响。研究2a-b证明了任务价值、人类身份差异和AI身份差异对大众对AI的感知的不同影响。研究3证明了任务类型对特定威胁感知的影响。研究建立了影响生成式AI接受度的推拉力模型和“人-AI-任务”影响框架，丰富了对生成式AI接受情况的认识，同时为生成式AI的合理应用及人机协作提供指导。

Generative AI is rapidly growing and its applications are controversial while receiving positive promotion. There are still few studies on the acceptance of generative AI. This study explores the core predictive roles of perceived threats and perceived competence on acceptance, as well as the effects of human identity, AI identity, task value and task type on perceptions and attitudes from the perspective of pull and push forces affecting acceptance, as well as the Human-Machine-Environment (HME) system. Study 1 demonstrated the effects of perceived threats and perceived competence on acceptance. Studies 2a-b demonstrated the differential impact of task value, human identity, and AI identity on perceptions of AI. Study 3 demonstrated the impact of task type on specific perceived threats. The study establishes the push-pull model and "human-AI-task" framework that affects the acceptance of generative AI, which enriches the understanding of the acceptance of generative AI, and provides guidance for the rational application of generative AI and human-AI collaboration.