Answer Provided by Galin Jones, Vice Provost for AI

The following question was recently posed, but similar versions are routine:

In your view, what does the University need to do to balance our embrace of AI with our interest in maintaining a culture of academic integrity?

I hope we continue to aspire to a culture of academic integrity and are serious about achieving it.  I am concerned about academic integrity in the classroom, as well as with research and other creative outputs.  

My experience with AI is that it exacerbates existing problems. AI tools have made concerns about academic integrity in the classroom demonstrably worse; the already-busy Office of Community Standards has been overwhelmed with reports of student academic misconduct, with no end in sight. While enforcement could be improved, I don’t believe it will, by itself, result in a culture of academic integrity. 

I also don’t believe that a blanket ban on AI tools or removing university-provided access to tools like Gemini will meaningfully address concerns about academic integrity, as these AI tools are already available to anyone with internet access who is willing to accept the provider's terms and conditions and, if applicable, pay for access. 

The policy for instructor and unit responsibilities is unchanged, and the university is not requiring AI in any instructional setting that I am aware of.  This recognizes that instructors have the expertise to decide whether AI should be avoided or required in their courses and degree programs.  However, the university would be irresponsible not to lead a conversation about where, how, and why it may make sense to use it.  This is part of what the AI Hub will do this year.

Concerns about academic integrity should not be limited to students and classrooms.  Research and creative outputs are also susceptible.  As expectations encourage increased “productivity,” some are already resorting to using AI tools. I won’t name this person, but a well-known faculty member at a top-ranked institution whose research interests overlap with mine has produced more than one paper per day so far in 2026.  (Note that increasing expectations have been happening for a long time, but, again, AI is exacerbating the situation.) Anecdotal reports abound about the deluge of AI-augmented papers being submitted to computer science and machine learning conferences, and about authors resorting to submitting slightly modified versions of the same paper in the hope of increasing their chances of success.  For the last several years I’ve been the editor for a scholarly statistics journal, and, seemingly overnight, the rate of submissions doubled and then tripled. Many of the papers are now obviously AI-generated.  Publisher policies are often permissive, only requiring acknowledgment of AI use.  

While I haven't provided concrete solutions to our shared concerns about academic integrity, I believe this is solvable.  In both cases, the key appears to be how work is assessed.   In the classroom, we need to be clear about pedagogical goals and the role assessment plays in achieving them.  In some classes, the method or process of assessment will have to change, while in others the content being assessed must change. In research, I suspect that these trends will continue until the ways we assess their value catch up.   This applies to journals, conferences, and other venues as well as for promotion, tenure, and merit.   

I’ll end by noting that the AI Hub is working closely with the Libraries, the Center for Educational Innovation, and Academic Technology Support Services to support instructors and units as we navigate the current situation.  The Teaching Support web page has a number of resources for instructors. We will also be rolling out several opportunities to engage throughout the academic year.