生成式人工智能对在线零工就业机会的影响——基于SSO框架与保护动机理论的实证研究

The Impact of Generative Artificial Intelligence on Online Gig Workers’ Employment Opportunities—An Empirical Study Based on the SSO Framework and Protection Motivation Theory

摘要: 生成式人工智能的规模化商用正在深刻重塑数字服务市场的生产结构与就业形态，在线平台从业者在政策语境中常被称作“灵活就业群体”，但劳动研究视角下属于不稳定就业群体，普遍存在劳动关系松散、社会保障缺失、收入波动大等特征，面对AI技术替代时抗风险能力薄弱，其技术压力下的心理适应机制亟待系统探究。现有研究多从宏观视角分析人工智能对就业总量的影响，缺乏针对零工群体这一特殊就业形态微观心理机制的系统性探讨。本研究基于应激源–应激–结果(SSO)框架与保护动机理论，构建人工智能影响在线零工就业机会的综合心理机制模型，通过对全国482名在线零工的有效问卷数据进行结构方程模型与Logistic回归分析，系统揭示技术冲击下零工群体的心理响应路径与行为应对逻辑。研究结果表明：任务替代风险与职业保护因素均通过感知危险与自我效能两条路径影响零工的应对行为，其中自我效能感对应对行为的驱动作用(β = 0.409, p < 0.001)显著强于感知危险(β = 0.327, p < 0.001)；人工智能使用比例(OR = 0.815, p = 0.027)与使用年限(OR = 0.865, p = 0.087)对替代担忧具有显著负向影响，而责任归属问题(OR = 1.313, p = 0.002)则显著加剧焦虑；应对行为对职业结果与未来预期的直接路径不显著(p = 0.113)，提示技术适应存在更复杂的传导机制，个体心理调整仅能缓冲外部结构性劳动风险，无法从根源消解AI带来的就业不稳定问题。研究结论为零工个体制定差异化技术适应策略、平台构建分层技能支持服务及政策制定者完善心理赋能公共服务体系提供了实证依据与决策参考。

Abstract: The large-scale commercialization of generative artificial intelligence is profoundly reshaping the production structure and employment patterns of digital service markets. Online gig workers are labeled “flexible employment” in official policy discourse, and are categorized as precarious work in labor studies. They generally have loose labor relations, lack social security, and large fluctuations in income. When facing AI technology substitution, they have weak anti-risk capabilities, and psychological adaptation mechanisms remain underexplored. Existing research predominantly examines AI’s impact on aggregate employment from macro perspectives, with limited attention devoted to the micro-level psychological mechanisms of gig workers—a distinctive employment form characterized by the absence of fixed labor relations and task-based compensation. Grounded in the Stressor-Strain-Outcome (SSO) framework and Protection Motivation Theory (PMT), this study constructs an integrated psychological mechanism model to elucidate how AI influences online gig workers’ employment opportunities. Through structural equation modeling and Logistic regression analyses of 482 valid questionnaires collected from nationwide online gig workers, this research systematically reveals the psychological response pathways and behavioral coping mechanisms under technological shocks. The findings indicate that both task substitution risk and occupational protective factors influence gig workers’ coping behaviors through two parallel pathways—perceived threat and self-efficacy—with self-efficacy demonstrating a significantly stronger driving effect (β = 0.409, p < 0.001) than perceived threat (β = 0.327, p < 0.001); AI usage proportion (OR = 0.815, p = 0.027) and duration of AI experience (OR = 0.865, p = 0.087) significantly mitigate substitution anxiety, whereas ambiguity in liability attribution (OR = 1.313, p = 0.002) substantially exacerbates concerns. Notably, the direct path from coping behavior to career outcomes and future expectations remains non-significant (p = 0.113), suggesting a more complex transmission mechanism in technological adaptation’s individual psychological adjustment can only ease structural precarious risks rather than fundamentally resolve employment instability caused by AI. These conclusions provide empirical evidence and decision-making references for gig workers to design diversified technological adaptation strategies, for platforms to launch tiered skill support services, and for policymakers to enhance psychological support within public service frameworks for flexible employment.