Kentucky requires AI 'proficiency' from all its college students, but skips the debate on jobs and energy

By Zendoric · September 5, 2026. The state task force that advises Kentucky's public universities on artificial intelligence (AI) —the Kentucky Council on Postsecondary Education (CPE) Task Force on AI, Higher Education and the Workforce— wants every college student in the state, whatever their degree, to be 'proficient' in the use of AI before graduating. The proposal came from task force member Jordan Harris at last Wednesday's meeting, on September 2, as reported by the Lexington Herald-Leader. The goal, he said, is not basic 'literacy' but 'proficiency': knowing how to automate research and problem-solving in any discipline, rather than leaving the matter to computer science departments alone.

The University of Kentucky (UK), the state's largest higher education institution, is already moving in that direction. Through its partnership with Microsoft —dubbed the Commonwealth AI Transdisciplinary Strategy (CATS AI)— it automates classroom learning modules and transcribes doctor-patient consultations across its hospital system, UK HealthCare. It also offers an introductory course, TEK 100, focused on 'collaborative intelligence' with AI. At the University of Louisville (UofL), executive director of AI strategy and innovation Jeff Guan went further: he said the first 'AI-native generation' has already reached the classroom —students who were in high school when AI broke through— and that faculty lag behind them in handling it.

But the report itself opens cracks in that narrative. The task force offered no evidence that students are the ones demanding more AI: according to a 2026 report by Gallup and the Lumina Foundation, only one in five U.S. college students uses it daily. Nor, according to the article, has the group discussed the environmental impact of the data centers that sustain AI —which consume large amounts of electricity and water, according to MIT— or potential job losses, even though studies from Stanford, from labor-market consultancy Challenger, Gray & Christmas and from the Economic Innovation Group already document, if in limited fashion, job cuts attributable to AI. It is also worth noting that Harris, the man calling for that across-the-board 'proficiency,' works as a client relationship executive at Deloitte Consulting, a firm that provides large-scale services to UK's own operations, technology and strategic projects: this implies nothing improper, but it does mean that some of those setting university AI policy in Kentucky have commercial interests aligned with its expansion.

Our take: Kentucky is a textbook case of how the transition to AI is being managed in U.S. public institutions, and not only in universities. The push comes from the top —state councils, consultancies, technology partners— rather than from real demand by students or faculty, and it is wrapped in the language of workforce readiness. The group's own internal debate gives it away: Rolf Provan, president of engineering firm Luckett-Farley and a task force member, asked whether Kentucky should 'lead,' 'catch up' or simply 'cover the bases' in the face of a Republican-majority state legislature; the council's executive vice president, Travis Powell, replied that it is sometimes wise to wait before taking major steps. It is a defensible logic: if the labor market is going to demand AI skills across every discipline —something we already see in sectors such as education, where the winning teacher is the one who knows how to orchestrate these tools rather than merely deliver content— universities had better teach it sooner rather than later. But mandating the tool without opening a debate about its real costs —displaced administrative jobs, energy consumption, who trains and pays the faculty who today lag behind their own students— is settling for half the equation.

That, at bottom, is the tension that defines this moment for AI in higher education. In the short term, mainstreaming its use without clear governance, adequate funding for teaching staff or an honest assessment of the impact will generate friction: overwhelmed professors, legitimate doubts about which administrative roles will survive, unanswered questions about energy spending that multiplies campus by campus. But the alternative —graduating generations who don't know how to work with these tools into a labor market that already demands them— would be worse in the medium and long run, because it would sideline precisely those who should benefit from the next wave of productivity. The challenge for Kentucky, and for any university system that sets out on this path, is not whether it should teach AI: it is whether it can do so with the resources, the faculty training and the honesty about side effects that, by this very report's account, it still lacks.