Artificial Intelligence, Teaching Quality and Management EfficiencyA Governance Perspective on Higher Education in Shenyang, China
Artificial intelligence may influence teaching and university administration, but the two areas do not create value in the same way. Treating deployment or processing speed as evidence of educational improvement risks overlooking weak learning outcomes, additional review work, bias, privacy concerns and blurred responsibility. Drawing on policy documents and a narrative reading of higher-education research, this article develops a governance framework for local institutions. The framework begins with a shared foundation of data, platforms, staff capability and institutional rules, then separates a teaching pathway from a management pathway. Teaching is considered in terms of learning goals, feedback, differentiated support, student engagement and teachers' professional judgement. Administration is considered in terms of turnaround time, error and rework, access to services, transparency and the cost of appeal. The analysis proposes an evidence cycle that moves from problem definition and risk classification to a limited pilot, human review, multidimensional evaluation and a decision to scale, revise or stop. For institutions in Shenyang, the practical priorities are shared technical standards with controlled data access, role-specific AI literacy, explicit evidence thresholds and an auditable division of responsibility between people and systems.