University of Tennessee trustees adopt systemwide policy on artificial intelligence for instruction

The University of Tennessee Board of Trustees approved a draft policy on the use of artificial intelligence for instructional and assignment purposes during its winter meeting.

The policy, presented by Dr. Matt Matthews, is intended to “set the guiding principles and expectations for our employees, faculty, and staff, and our students around the use of artificial intelligence for instructional and assignment purposes,” Matthews said. He told trustees the policy responds to a Tennessee law that requires each public university to adopt an AI policy and that it draws on federal guidance, including President Biden’s Oct. 30, 2023, executive order on AI.

The draft defines artificial intelligence for the board using language drawn from statute and the executive order and notes common capabilities of generative AI tools such as text and image synthesis, summarization, and automation of repetitive tasks. The policy frames AI as a tool of potential educational benefit while preserving faculty authority over course expectations and upholding academic integrity and data-protection rules.

The policy requires each campus to publish clear, course-level statements about what kinds of AI use are allowed or prohibited in a given class and says that violations should be handled under each campus’s academic integrity procedures. Matthews also told the board that AI-based detection tools may be used to flag possible AI involvement but “they cannot be the only tool used to determine whether a student has engaged in misconduct.”

Members of the board and attendees who spoke during the discussion urged the policy be flexible enough to adapt as AI tools evolve and asked for periodic review. Trustee Charlie (last name not specified in the transcript) asked whether the policy could “blow up”; Matthews and other speakers responded that the system has the ability to revise the policy as needed and that campuses will implement procedures appropriate to their programs and disciplines. A student who identified herself as Lisa described most professors as already including AI statements in syllabi and said the policy would make that practice systemwide.

A motion to approve the proposed policy was made and seconded; the chair called for a voice vote and announced the motion approved. The transcript does not record numeric vote counts.

Trustees and staff said they expect campuses to implement course-level procedures quickly and to return to the board if significant changes to the system policy are needed as tools and practices evolve.

Matters noted but not changed by the vote included: (1) the requirement that course statements be clear about allowed AI tools and uses; (2) faculty retain prerogative to set disciplinary standards for original student work; and (3) use of AI detection software is permitted as an investigative aid but must be supported by disciplinary review rather than used alone to establish misconduct.

The board moved on to other business after the vote.