Who Should Teach AI? The Questions Universities Need to Ask

Artificial intelligence is now a flagship subject. Universities across the UK are launching AI degrees, AI modules and AI research centres, and students are enrolling in record numbers to learn how to build with it. That enthusiasm deserves better than it sometimes gets, and it raises a question that institutions should be brave enough to ask: who is actually at the front of the room?

In my opinion, the academics at a university are supposed to be teaching artificial intelligence to the people who are now learning it: students who want to build genuine AI knowledge, and who will go on to engineer real systems. If that is the promise, then important questions need to be asked.

Four questions worth asking

Not as an interrogation, but as due diligence. Before an academic teaches AI engineering, it is reasonable to ask:

1. What projects has this academic undertaken to gain this experience? Where is the work that produced the understanding they now teach?
2. Have they developed a system of significance? Something with real users, real constraints and real failures: evidence that the knowledge was earned, not merely acquired.
3. Can they explain, academically and practically, how an LLM works, and how it actually helps software engineers who are working or planning?
4. Are they a forward-deployed engineer? Hands-on delivery experience, including multi-agent orchestration: the realities of building systems, not just describing them.

Questions are important. Questions and evidence.

Credentials tell you someone attended. Evidence tells you someone built. A university promising AI engineering should be able to show the second.

Why this matters to institutions, not just students

Once the important answers are in, once there is evidence to back up an academic's claimed skills and knowledge, you will see progress in the academic industry, especially at universities. The reverse is equally true. If a course becomes known for teaching that is questionable, word spreads. Students talk, applicants drift elsewhere, employers learn which graduates can actually build, and a reputation that took decades to build can be damaged in a few intakes.

A university is a brand. Its teaching is its product. In a market where institutions compete globally for the same students, "who teaches this?" is not a rude question; it is the question.

The uncomfortable scenario

Consider the case of an "academic" teaching a subject they claim to understand and be an expert in. Have you really made sure your assumptions are correct? Or is the teaching a cover for professional experience that could be fabricated: a career talked through with words and certificates rather than built with systems?

If those assumptions and answers are verified properly, you eliminate the fear of bad teaching for future students and enable a positive, impactful experience going forward. If they are not, the students pay the price, and eventually, so does the institution.

Holding a higher bar: the reward for meeting it

This is not an argument against academics. It is an argument for evidence. Many excellent academics combine rigorous research with genuine delivery experience, and they are the standard worth hiring, supporting and promoting. The goal is simply to make that the norm rather than the exception: a subject taught by people who have mastered it, in classrooms where practice and theory reinforce each other.

Think about the future of the students: the future they will build with your teaching. They deserve instructors whose experience is real, and institutions deserve reputations that are earned. Ask the questions. Demand the evidence. Then build.

If you're a university or training provider reviewing an AI curriculum and want an honest industry view, email brandon@kreostudio.co.uk.