UBC’s approach to academic integrity in the age of AI

This story was not written by ChatGPT. Or was it?

This story was not written by ChatGPT. Or was it?

It’s the kind of question many university professors are now asking themselves as they seek to evaluate student work while the capacity and capability of artificial intelligence systems such as ChatGPT expands almost on a daily basis.

“UBC is well aware of generative AI tools such as ChatGPT and their potential use in academics,” said Simon Bates, Vice-Provost and Associate Vice-President, Teaching and Learning pro-tem.

“Our primary focus at UBC around academic integrity is an educative approach: to help students and faculty understand better the importance of academic integrity and provide them with resources to not only reinforce that importance, but also to enable them to understand what constitutes academic misconduct, and how to avoid it,” said Bates.

The university has convened advisory groups on both the Vancouver and Okanagan campuses with representation from faculty, students and staff to consider how it can do more on the educative side of academic integrity: to look at how different course designs might be used to support academic integrity, how better to define and communicate to students what is and is not acceptable in their various course contexts, and to provide resources and encouragement to instructors to discuss the importance of AI in their classes.

“A key output from these advisory groups has been resources for faculty and students around current issues in academic integrity, such as the role of AI tools and services. For example, shortly after the public release of ChatGPT v3, we collated reports, guidance and advice for faculty ahead of the start of this semester as they finalized course syllabi. We developed a newsletter and FAQ as part of our educative approach to academic integrity, which was shared with faculty and students,” said Bates.

As impressive as the current iteration of these tools are, they have significant limitations. ChatGPT is currently trained on data up to 2021, so cannot generate context on events beyond that time, and has been shown to generate incorrect responses and invalid references. Tools to assess the authors of written text – human or AI – are also rapidly developing.

While some institutions are “banning” programs like ChatGPT, Bates said that’s not an approach being considered right now at UBC.

“Banning a tool like this is unlikely to have the desired goal of reducing its use as a way to subvert academic honesty. As with many emerging tools and technologies, generative AI technology comes with both potential benefits and real challenges; it has the potential to support and enhance learning, but can also be used to pass work off as your own. Faculty members and their departments are best placed to decide if and how to make use of such tools in their curricula, and / or as tools to support student writing. These discussions are actively on-going,” he said.

Academic integrity is fundamental to UBC’s core mission of as an institution of higher learning and research. As a globally ranked university, UBC holds its students to a high-level of conduct and expects them to be honest and responsible members of the academic community. The university conducts thorough investigations of all allegations of academic misconduct and takes disciplinary actions when warranted. An annual report containing summaries of student discipline cases is published here.