生成式人工智能正在推动高等教育从传统治理、数字治理向生成式治理演变。这一治理是由大模型、大算力与大算据驱动衍生的全新范式，可通过人机协同、人机互动与场景训练达成前瞻决策、动态响应、调节反馈与协同决策的生成涌现功能，为高校战略规划、教学管理与学科调整提供科学决策支持。高等教育生成式治理起始于知识涌现机制，经由自组织进化与价值对齐过程而不断趋于完善，但亦面临价值约束不足、基础支撑薄弱、协同进化梗阻与场景转化困难等实践困境。面向未来，高等教育需从“伦理—环境—能力—评价”四个维度构建“人在领路”的治理共同体，推动全域数据融通、整合多元优势与创生多场景治理策略，从而彰显人机协同治理决策的生成主义价值。

Generative artificial intelligence is driving the evolution of higher education from traditional governance and digital governance to generative governance. This governance model is a new paradigm driven by large models, high computing power, and massive data. It can generate and emerge in the form of forward-looking decision-making, dynamic response, regulatory feedback, and collaborative decision-making through human-computer collaboration, human-computer interaction, and scenario training, providing scientific decision support for university strategic planning, teaching management, and discipline adjustment. Generative governance in higher education begins with knowledge emergence mechanisms and is constantly improving through self-organized evolution and value alignment processes. However, it also faces practical challenges such as insufficient value constraints, weak basic support, obstacles to co-evolution, and difficulties in scenario transformation. Looking to the future, higher education needs to build a human-led governance community from four dimensions: ethics, environment, competence, and evaluation. This will promote the integration of data across the entire domain, consolidate diverse advantages, and create multi-scenario governance strategies, thereby highlighting the generative value of human-machine collaborative governance decision-making.

G64；G434

国家社会科学基金项目（CIA240281）