Skip to content

CU to consider new policy on AI

Policy comes at CU prepares to roll out campus-wide ChatGPT on March 31

Author
PUBLISHED: | UPDATED:
Getting your Trinity Audio player ready...

The University of Colorado is developing a new policy surrounding artificial intelligence as it prepares to roll out a CU-controlled version of ChatGPT effective March 31.

The proposed policy recognizes that AI “presents ethical, legal, and operational challenges” to the university, the draft read. The policy states that the university will use AI appropriately while ensuring privacy, security, transparency, fairness and human oversight. The University of Colorado Board of Regents discussed the proposed new policy in a governance committee meeting on Thursday.

The proposed policy, which is a regent policy, will be paired with an Administrative Policy Statement on AI. Regent policy is a high-level policy that directs Administrative Policy Statements, or APS. The regent policy is a brief, high-level overview, while the APS will contain all the details of CU’s AI policy.

The Administrative Policy Statement on AI is about five pages long, CU System Chief Information Officer Scott Munson said, and establishes campus and system oversight structures for AI by working with shared governance groups. Shared governance, including groups such as the Boulder Faculty Assembly and Staff Council, is a collaborative model that allows faculty and staff to have a say in decision-making.

“It’s going to set the guardrails around responsible use of GenAI,” Munson said.

GenAI, or Generative AI, is a type of artificial intelligence that creates content from existing data.

The CU System and Board of Regents are trying to complete and establish the policy promptly because of the ChatGPT rollout that’s happening, according to CU System Chief of Staff Leonard Dinegar. CU has entered into a $2.1 million contract with OpenAI, providing ChatGPT Edu environments to all CU campuses starting on March 31. Dinegar acknowledged the concerns people have shared regarding AI at CU, and said the university is working to mitigate those concerns. Students, faculty and staff have expressed concerns such as privacy, academic misuse and cost.

An initial draft of the AI Administrative Policy Statement is posted on CU’s website, but has since changed significantly due to public feedback, Munson said. A new draft will be posted by CU’s Office of Policy and Efficiency in the coming weeks and will be available by visiting cu.edu/ope/aps/under-review.

Faculty Council Chair Jorge Chavez said CU’s administration has been largely responsive to faculty concerns and collaborative throughout the policy-creation process.

“I know we’re trying to move quickly, given the necessity, but we want to be really careful about making sure we’re addressing all concerns that are being raised,” Chavez said. “It’s definitely a work in progress, but we are making good progress.”

The policy was first introduced in a regent committee meeting on Jan. 23 and will go to the full board for approval at its next regular meeting on April 16 at CU Denver.

Subscribe to our new newsletter to get CU Boulder’s latest news sent straight to your inbox.

RevContent Feed