AI in Higher Education: Overcoming Challenges and Building the 'Competent Institution'
Reacting Beyond the 'Status Quo'
With this "deteriorating" projection ahead, there is a strong argument to be made about some type of reaction. On the one hand, the trust and sense of value in higher education needs to be restored, and on the other hand, and in parallel, significant improvements need to be made to stop the current and projected financial situation from getting even worse. An oversimplified view I have of these two elements is that institutions need to improve and/or innovate their product, meaning their academic offerings, as well as improve their operating efficiency. This is a very tactical view, and it is so purposely, so that it is explicitly understood that this reaction needs to go beyond shelved and inactionable strategic plans, endless academic discussions, SWOT analyses, and reports of all kinds. Having had the fortune in my career of existing, often at the same time, in the higher education strategic and tactical worlds, I've had a long-held view that at some point "we need to land the plane." I use this metaphor often and in a very colloquial way to signify that at some point for every plan, for every analysis, for every discussion, there needs to be a "finish line" of sorts that needs to be crossed, and some element of tangibility needs to be attained, even at the risk or expense of failure. In previous years I have proposed such tactical approaches in "
An Inflection Point for the Creation of New Cybersecurity Operating Models in Higher Education" and "
Avoiding the College Enrollment Cliff With AI."
Particularly as it relates to AI, because of the roles I've had for the past four years, coupled with my academic formation, and the years I was a CTO in higher education, my views on the potential positive impact of AI have expanded significantly. I have not only studied AI, I have also deployed it and advised others on it. Hinged on this experience, I am a very strong proponent that AI can significantly improve institutional and operational efficiency. Unfortunately, and perhaps customarily, the higher education segment, despite the challenges and pressures explored earlier in this article, is not funding, adopting and deploying AI for operational efficiency nearly at the rate others are. In a recent 2024 study (Now Decides Next: Insights from the Leading Edge of Generative AI Adoption), Deloitte found that 91% of surveyed organizations expect their productivity to increase because of the adoption of AI. The same study found that 42% of the surveyed organizations are already reporting efficiency, productivity, and cost reduction being achieved with the use of AI. This is very tangible; this means that almost half of organizations in other segments are attaining that tangibility I mentioned earlier. Conversely, a survey conducted by Forbes ("Higher Ed Leadership Is Excited About AI — But Investment Is Lacking") found that only 21% of surveyed institutions believe they are prepared to fund and deploy AI operationally.
The Competent Institution and AI
Educause in its more recent
report of top 10 priorities for 2025 organizes higher education top priorities in groups, and in doing so it defines the "Competent Institution." This group of priorities is all about institutional efficiency, agility, streamlining, and modernizing both process and infrastructure. This "Competent Institution" grouping of priorities is without a doubt a very fertile ground to plant seeds for efficiency improvement. By direct correlation, this in my opinion is where AI thrives. In writing the final part of this article I want to provide some measure of clarity around the topic of AI use cases and their impact on efficiency gains. By far, the most discussed and deployed AI use case is the one of a digital assistant. With the advent of generative AI, and the ease with which retrieval augment generation (RAG) can be done, it is not surprising its deployment is so prevalent. Having seen the case operationally in many settings, I can surely say it helps to increase workforce productivity. These digital assistants also have an effect of improving customer (student/faculty/staff) experience, which counts toward institutional success. These productivity gains sometimes result in efficiency gains as well. I however feel that the efficiency gains from digital assistants in higher education might not provide a significant enough contribution to help change the current trajectory of many institutions at risk. The more significant efficiency gains may come from other applications of AI, for example (not exhaustively):
- Optimization of physical plant utilization.
- Implementation of real-time/near-real-time processing to support the recruitment/enrollment/graduation journey from prospect through matriculated student, and eventually through graduation.
- The use of sensors and edge compute/AI to optimize how services are rendered in cafeterias, parking lots, tutoring and counseling centers, libraries etc., to optimize service provisioning and with that user experience, as well as to minimize waste.
- Optimization of class scheduling to adjust the ratio between the number of sections and number of students per section, as well as the number of classroom and buildings in use at any given time.
- Personalization of marketing efforts and fundraising activity to create a more unique experience for the constituent groups being serviced, while enhancing, and really optimizing internal operational efficiencies.
Any one of the previously mentioned examples, and there are more, may require significant planning and investment in some cases, however, the results are directly proportional to the level of effort. The key to all this is to start somewhere. The clock is ticking.
About the Author
Hernan Londono is chief technology & innovation strategist, higher education, at Lenovo.