The use of generative AI is on the rise in college and university classrooms, despite lingering concerns about its potential for bias, cheating, inaccuracy and other risks. That’s a main takeaway from a new report by Cengage, entitled At the Trailhead: The State and Future of AI in Higher Education.
The report is based on a survey of 967 instructors who have taught college courses for one or more years. A similar survey was conducted last year, and a comparison of the results shows a notable increase in the use and positive perceptions of GenAI by college instructors.
Here are some of the highlights:
In one year, the use of GenAI has nearly doubled among college instructors, from 24% in 2023, to 45% in 2024, and just under half (49%) of instructors now have a positive view of generative AI, up from only 28% in 2023.
Almost two-thirds (63%) of college instructors believe AI can improve student engagement, and 71% believe AI can support lifelong learning.
Nearly all (90%) of the surveyed instructors believe that generative AI technology will play an increasingly important role in education in the coming years, up from 73% who agreed with that statement last year.
In addition to those college faculty already using GenAI in their classrooms, 41% said they planned to use it in the future compared to only 15% who said they had no plans to incorporate Gen AI into their classrooms.
Instructors endorsed a wide range of current and potential uses for GenAI, including providing personalized student learning, assisting with lesson planning, simplifying various administrative tasks, promoting better classroom engagement, and conducting assessments.
Concerns Remain
Despite this enthusiasm, 89% of college instructors remain worried about various ethical concerns, privacy problems and other risks of GenAI. For example,
- 82% of college instructors are concerned about matters of academic integrity such as cheating and plagiarism;
- 50% are concerned about the accuracy of AI-generated outcomes;
- 40% cite concerns about the potential for bias with GenAI;
- 34% are concerned about the risks of manipulation or harmful use;
- 29% are concerned about data privacy;
- 28% are concerned about the ethics of obtaining training data;
- 26% are concerned about the possibility of deskilling;
- 23% of higher ed instructors say there is a need for more training and support around AI adoption;
- 16% expressed concern about job loss due to AI’s growth;
- 14% are concerned about cybersecurity risks.
The contrast between perceived risks and rewards is captured by these two quotes from survey respondents:
“When students rely on AI instead of learning to writeand study, they learn less. This is like giving kids a calculator before they know what it even means to add and subtract. This is not a positive development for learning,” said one adjunct professor.
“Whether I like it or not, it is already impacting our daily lives, and its impact will only get greater. I believe it is important for us to wisely harness its potential while carefully reducing its potential harms,” observed an assistant professor from Arizona.