Universities weigh costs and rules as they bring AI into teaching

In short: Universities are adopting AI in very different ways, and money, training needs, and student concerns are shaping what they choose.

What's going on

Universities around the world are trying to decide how to use AI tools in teaching and learning. Some want to teach students how to use AI, while others are moving more slowly because the tools can be expensive and still feel untested.

Ohio State University says it will embed AI across its courses so that undergraduates graduating in 2029 are “AI fluent.” Other schools are narrowing their plans because they also need to pay for staff training and student training, at a time when many universities face tight budgets.

Cost and access issues are showing up most clearly outside richer countries. Leaders in Indonesia and the Philippines said many students rely on free versions of AI tools, which can limit what they can do. In parts of Sub-Saharan Africa, some campuses do not have reliable internet, which makes online AI tools harder to use, like trying to stream a video on a weak signal.

Universities also face mixed attitudes and skill levels. Some students arrive knowing more about AI than their lecturers, while others avoid it because of worries about bias (when a system treats groups unfairly), data privacy, made-up answers (often called “hallucinations”), environmental impact, and being accused of cheating. At Queen Mary University of London, a student union study found about 20 percent of respondents did not use AI at all.

What to watch

More governments and regulators are starting to respond. Australia’s higher education regulator has issued guidance that includes training staff, and Singapore has set up a committee to coordinate approaches. Expect more shared rules on what is allowed in coursework, and more interest in smaller, locally run AI systems that can be cheaper and easier to control.

Source: Financial Times