The Carleton University Senate met on Sept. 22 to discuss the challenges of generative artificial intelligence (AI) and considered recommendations for its integration in an academic setting
Generative AI refers to algorithms capable of creating text, images or media. Alongside other post-secondary institutions, Carleton University is exploring the use of AI technology in education, which previously posed academic integrity concerns.
The working group on the use of AI in teaching and learning at Carleton University proposed their report at the meeting. The group consisted of Carleton instructors, students and staff aiming to assess AI tools in education as new research becomes available.
The group will meet twice per academic year as generative AI evolves, according to David Hornsby, associate academic vice-president at Carleton and chair of the working group,
“This report offers no silver bullets,” Hornsby said. “There is not one or two or three solutions immediately available to take advantage of the opportunities or address the academic integrity challenges.”
He said the working group seeks to evaluate AI’s potential rather than restrict its use.
While the report acknowledged the Academic Integrity Council of Ontario’s statement that AI may allow students to offload academic work, it recommended university instructors “chart innovative ways to utilize [AI] to support key student learning outcomes.”
Hornby proposed nine guidelines about the use of generative AI to be implemented into Carleton’s Academic Integrity Policy.
The working group also recommended several short and long-term strategies for academic integrity, from proctoring and oral exams to encouraging students to develop their AI literacy.
Hornsby said he hopes these strategies are implemented into the Academic Integrity Policy. In the meantime, he aims to “circulate [the report] more broadly to the community as soon as possible.”
“This is a time for creativity while not losing key learning opportunities, such as learning to write efficiently and beautifully,” he added.
Pamela Wolff, a chemistry instructor at Carleton and member of the Senate, said she hopes to see AI detection tools used to inform—rather than discipline—students.
“I would like my own students to run their work through [AI detectors], and provide commentary on/about whether they are writing like a chat bot,” Wolff said.
However, commercial AI detector tools are remarkably inaccurate and often monetized, Hornsby said.
Early adopters of revised AI policies have seen unprecedented success. Wolff, who recommends circumventing rather than seeking out academic integrity violations, saw generative AI appeals drop 90 per cent.
“Following COVID-19 restrictions, cases of AI have drastically decreased and especially relating to final exams,” said Cristina Ruiz Martin, the appeal committee chair.
A report by the Senate Student Academic Integrity Appeal Committee showed additional decreases in Academic Integrity Policy violations across all undergraduate faculties since July 2022.
The Senate is scheduled to meet again on Oct. 20.
Featured graphic by Sara Mizannojehdehi.
