What Universities Need to Be Doing in 2025
If you think AI is something your IT team can “get to later,” think again. The future of your institution doesn’t hinge on the next CRM upgrade or a shinier LMS, it hinges on whether you can take control of your data, your AI strategy, and your institutional talent before someone else does.
Let me explain.
A few weeks ago, I was shopping for a new laptop. My laptop battery started overheating so I decided it was time to buy a new one. My requirements were simple enough: Linux-friendly, large SSD, solid GPU. During my search, I realized something. I noticed graphics cards are important for AI developers. You can choose from different levels of graphics cards depending on your needs:
- No AI Development
- Budget AI Dev: coding, inference, basic training
- Mid-tier AI Work: good for training, diffusion
- Heavy AI Training: model finetuning, Large Language Models (LLMs)
This led me to a question. Don’t we just log in to online servers and train/run AI models there? Who needs to train LLM’s locally? Surely, we just use them; maybe giving them a context prompt.
Well, powerful GPUs aren’t just for gamers anymore. They’re the backbone of AI work. And while I wasn’t planning to train a Large Language Model in my living room, it struck me that this isn’t going to stay optional. Training AI will soon be fundamental. Just as every company trains its salesforce, every organization, including universities, will need to train and run their own AI models. Why? Because chatbots, not humans, are already your biggest audience. Prospective students, alumni, faculty will all interact with your institution through AI long before they read your website.
So let’s get blunt: universities cannot afford to sit back. CIOs need to lead their institutions on three fronts: owning their data, leveraging AI, and building the right talent base before higher ed becomes the next industry disrupted beyond recognition.
- Own Your Data
If you don’t own your data, you don’t own your university’s future.
This isn’t a technology problem. APIs, connectors, integrations, they exist, and they work. The problem is business models. Vendors profit by keeping your data captive, and too many universities are comfortable with that tradeoff.
But AI raises the stakes. If your institutional data lives in a walled garden, you can’t connect it to new AI tools, and your competitors will leap ahead. Worse, you risk being locked into biased, opaque recommendations from someone else’s LLM.
Owning your data is no longer about “better reporting.” It’s about institutional sovereignty. If you don’t own your data, you don’t own your university’s future.
- Leverage AI (Before It Leverages You)
AI won’t wait for your governance committee.
Let’s be clear: AI isn’t replacing your staff; it’s amplifying them. The newest models can pass every PhD exam you can throw at them. Hand those tools to your faculty and administrators, and suddenly every employee is working with the brainpower of a thousand PhDs.
The challenge isn’t whether AI will make you smarter. It’s whether your systems can adapt at the speed of change. Agile methods, continuous deployment, automated testing, things IT leaders once called “best practices” are now survival skills.
The question isn’t whether AI will make your institution smarter. It’s whether your institution and its systems can adapt fast enough to keep up. Agile practices, automated deployments, robust testing, all the things your IT department has been “meaning to get to” are now survival skills. If you can’t deploy at speed, your AI advantage will evaporate.
- Secure the Right Talent (Because the War for Skills Is Real)
Yes, the headlines are full of layoffs in tech. No, that doesn’t mean skilled staff are easy to find. The real crisis isn’t too few jobs, it’s too few people with the expertise to integrate AI into business systems that actually work.
AI will automate tasks, sure, but that only increases demand for technologists who can guide, govern, and integrate it. For universities, that means a hiring and retention strategy that values not just AI experts, but the architects who can keep your systems future-ready.
The CIO’s Checklist for 2025
Here’s the uncomfortable truth: higher ed institutions don’t get a grace period. AI is already reshaping industries, and universities that hesitate will lose ground fast.
Ask yourself:
✅ Is your data free or trapped?
✅ Are your systems agile or sluggish?
✅ Do you have the talent to integrate AI or are you hoping vendors will solve it for you?
The clock is ticking. The universities that act today will define tomorrow. The rest will wonder how they lost control of their future.
Guest author: Robert Metcalf
To hear insights from other Higher Ed leaders, listen to this on-demand webinar:
Transforming Higher Ed through AI: Perspectives from Presidents, Provosts, and CIOs