The elephant in the classroom

This week marks the opening of the semester at UP Cebu where I teach media and communication law and ethics. I plan to conduct my first class meeting by addressing the elephant in the room head-on. That elephant, of course, is artificial intelligence or AI.

Early last week, the UP College of Law published its policy on AI and technology use in its first-year Juris Doctor courses. I agree with its position on AI: neither a wholesale rejection of the technology nor an uncritical adoption of it.

The College acknowledges that lawyers will inevitably work in a profession increasingly shaped by AI and that students must eventually learn to use these technologies ethically. But it draws an important distinction between becoming AI-literate and becoming AI-dependent.

AI literacy refers to “the knowledge, skills, and attitudes needed to understand how AI works, critically evaluate its outputs, and use it ethically, responsibly, and creatively.” This is consistent with the 2026 formulation of the Organisation for Economic Co-operation and Development (OECD) and with UNESCO’s AI competency frameworks.

AI dependency, on the other hand, describes excessive reliance on AI in ways that can weaken independent judgment or skills. In its 2024 generative AI risk-management framework, the U.S. National Institute of Standards and Technology (NIST) identifies “over-reliance” on AI as one of the risks arising from human interaction with the technology.

The challenge, then, is not whether to use AI, but how to use it without surrendering our own judgment to it.

In my opening lecture later this week, I will tell my students that AI is already the elephant in the classroom. Pretending that it does not affect virtually every aspect of university learning and pedagogy would be deeply disingenuous.

I use AI myself, largely as an assistant, because a professional who refuses to understand a technology that is reshaping their field will soon be unable to teach, criticize, or regulate it intelligently. I am, therefore, trying to be AI-literate. If a tractor can do in a morning what takes a carabao a week, why insist on the carabao?

But here’s the thing. AI can never do the thinking for you, just as a tractor does not know which field to plow. A farmer must first learn the land, just as a law student must read cases and study the course materials. Otherwise, the student would be unable to independently distinguish correct information from AI hallucinations or AI slop.

There is also the danger of cognitive debt, which is getting the task done today by outsourcing the thinking that should have trained your own mind. That matters in law and journalism. Those who have substance between their ears can instantly draw on it when faced with oral arguments, cross-examinations, interviews with sources, negotiations, examinations, and moments when a client, judge, or editor expects an answer without one having to look at a gadget or read from AI-enabled glasses.

We are biological beings. Our brain runs on energy derived from the calories in the food we eat. AI, on the other hand, is a machine that requires enormous amounts of electrical energy to process information or “think.” When we stop exercising certain mental capacities and routinely let AI do the thinking for us, we risk weakening those capacities. At least when a muscle weakens, we can see it. But when our capacity to recall, reason, write, argue, or make judgments begins to weaken, we may miss the signs until it is too late.

Before we teach our students how to think with AI, we must first make sure they understand the importance of thinking without it.