The Use of Artificial Intelligence in Higher Vocational Colleges in Sichuan, China: A Mixed-Methods Study of Adoption, Utilization, and Policy Implications

Liu Yang^1,2*Francis G. Balazon^3*

• ¹Urban Vocational College of Sichuan, Chengdu, China
• ²Urban Vocational Colleges of Sichuan, Chengdu, Sichuan, China
• ³Batangas State University, Batangas City, Philippines

This study investigated the status of Artificial Intelligence (AI) adoption and utilization in higher vocational colleges in Sichuan Province, China, and examined associated challenges and policy implications. A mixed-methods, descriptive-comparative approach was used, combining a survey of 1,085 respondents (administrators, teachers, and students) with interviews and observations. The aim was to assess current AI adoption (including prevalent AI tools, usage frequency, and purposes), evaluate the extent of AI integration in teaching, language learning, and research, compare perceptions across stakeholder groups, and identify key issues hindering AI use. AI adoption in Sichuan’s vocational colleges was moderate. Approximately 70–75% of faculty and administrators had adopted AI tools, compared to a lower rate among students, of whom nearly one-quarter were unsure if they had used AI. AI was used more for routine tasks and language learning activities than for research. Language learning emerged as the most AI-integrated domain, with widespread use of AI-driven translation and tutoring applications, whereas AI utilization for research purposes was relatively limited. Domestic AI platforms (e.g., Baidu ERNIE Bot and Alibaba’s Qwen) were the primary tools employed, reflecting China’s tech ecosystem and access restrictions, while foreign models like ChatGPT saw minimal use. Notably, only a small minority (under 10%) of respondents had received formal training in AI use, underscoring insufficient institutional support. Statistical analyses found no significant differences in overall AI utilization levels among students, teachers, and administrators, suggesting a cohesive adoption pattern across roles. However, those individuals who had prior exposure to AI training or institutional support exhibited significantly higher usage rates. Key challenges identified included limited training opportunities, inadequate infrastructure and access, uncertainty or resistance in using AI, and concerns regarding data privacy and academic integrity. In response, the study developed a comprehensive academic policy framework to guide ethical and effective AI integration across teaching, learning, research, and administration. This includes recommendations for capacity-building (faculty and student AI literacy programs), infrastructure enhancement, clear usage guidelines, and ongoing support. The findings and proposed policy provide actionable insights for educational leaders to harness AI in vocational education responsibly, thereby aligning with China’s national strategies for educational modernization.