WHAT remains if academic criteria shift from personal understanding to better prompt-driven outcomes generated by artificial intelligence (AI)? A lecturer associated with the University of Punjab recently highlighted this concern in a social media post, lamenting that many students who submitted well-written assignments often fail to explain their own work during viva examinations. This reflects a broader shift where academic performance increasingly depends on AI rather than genuine understanding.

Easy access to such a catalyst as AI in the hands of students is unarguably affecting students’ ability to think, argue, question, analyse and apply knowledge. This is making the educational structure a fragile framework, with students having a hollow conceptual base. This limits skill develop-ment, and risks producing a workforce that can operate tools, but cannot trouble-shoot the underlying logic when those tools fail.

It is not necessarily a sensible solution to exclude AI altogether when the world has widely recognised its unmatched convenience. However, carefully considered boundaries must be drawn to cultivate students’ potential.

Furthermore, academia should introduce evaluation methods that require direct intellectual engagement, such as oral examinations, in-class problem-solving and supervised assessments, that ensure education measures understanding, not just the ability to write prompts.

Muskaan Tariq
Karachi

Published in Dawn, July 13th, 2026