Fifty Systems, One Pilot: State Policy and Faculty GenAI Adoption in Tennessee Higher Education

• In Event: The Emerging Role of AI in Education and Workforce Preparation

Abstract

As the federal role in higher education AI policy recedes, US state systems are becoming the primary locus of decisions about generative AI adoption in colleges and universities. State contexts vary in workforce framing, system-level guidance, and equity mandates, and this variation governs the conditions under which faculty integrate generative AI into teaching. Yet policy scholarship on AI in higher education has largely treated adoption as a federal or institutional phenomenon, overlooking the state as a unit of analysis. This gap matters because faculty adoption is where state policy meets classroom practice, and where students experience the downstream consequences of state-level decisions.

This paper makes two contributions. First, it presents a comparative policy analysis of state-level higher education AI policy across the fifty US states. Drawing on state legislation, executive actions, public university system guidance, and state coordinating board documents from 2022 through 2026, the analysis constructs a typology of state postures toward generative AI in higher education along three dimensions: workforce framing, system-level governance guidance, and regulatory or legislative intervention governing academic freedom and curricular decisions. The typology establishes comparative infrastructure the field currently lacks and identifies clusters of states whose policy environments produce distinct conditions for faculty AI adoption.

Second, the paper grounds this typology in the Tennessee case through an ongoing institutional pilot of the Higher-Order Thinking with Generative AI (HOT) instructional model at the University of Tennessee, Knoxville, supported by $221,000 in internal funding from the Office of Research, Innovation, and Economic Development and the College of Education, Health, and Human Sciences. The HOT model, which received the 2025 AAAI Education Innovation Award, focuses on faculty and student use of generative AI for higher-order in place of substitutive tasks. We position the pilot within Tennessee's state policy environment: a workforce development orientation that treats higher education as labor market preparation, a public university system without an explicit AI-in-education framework, and a legislative context shaping academic freedom. This framing is consequential because it informs which AI competencies the state prioritizes for students entering a fast-changing labor market.

A prepilot phase convened 12 faculty and administrators in a workshop that surfaced readiness, concerns, and policy-adjacent tensions, and informed the scaled design. The main pilot is now in its second of four semesters, scaling the HOT model to 12 faculty reaching approximately 350 students through 2028. Mixed-methods data collection tracks instructional implementation, student outcomes, and faculty interpretation of state and institutional guidance. The conference paper will present the full policy analysis and typology, the Tennessee case located within it, and implementation findings from the first two semesters.

We argue that state policy scholars studying AI in higher education need a comparative state-systems lens. Tennessee, read against states with explicit AI-in-education guidance and states with restrictive postures, illustrates how state policy variation produces predictable on-the-ground differences in faculty adoption. The paper concludes with implications for policy design, system governance, workforce strategy, and comparative research infrastructure across the fifty systems.

Author

• Fatima Tuz Zahra, University of Tennessee at Knoxville
  Presenting Author