When AI Does Students’ Assignments, Are They Still Really Learning?

The arrival of artificial intelligence (AI) has changed many aspects of education. Tasks that once took hours — finding references, drafting writing, summarising reading and preparing presentation materials — can now be completed in minutes, or even seconds. With a simple instruction or prompt, AI can produce polished text, create presentation slides and answer questions that would previously have required a lengthy process of thought.

For students, AI has become highly familiar. It is not only a learning aid, but also a companion for completing assignments, preparing presentations and finding answers when responding to questions in class. This convenience cannot be ignored.

AI can help students understand course material they have not yet mastered, find additional information and improve writing they have produced themselves. But alongside these benefits, one question is increasingly troubling me: when AI does students’ assignments, are they still really learning?

This question does not come from a rejection of technology. I am well aware that AI is part of progress and cannot be avoided. The problem arises when this convenience leads students to skip the learning process itself. As a lecturer, I feel concerned, even saddened, to see AI increasingly dominating students’ academic activities.

Assignments, which should provide an opportunity to process knowledge, construct arguments and develop thinking skills, have instead become an exercise in transferring AI-generated output into students’ files. Presentations that should reflect students’ understanding can also be put together quickly using AI.

This dependence is even apparent during presentations and classroom discussions. It is not unusual for students to hold their phones and read from them while presenting material. During question-and-answer sessions, students asking questions sometimes read questions they have prepared with AI on their phones. Likewise, when answering questions from lecturers or classmates, they turn to AI again to formulate a response.

This situation leaves me wondering how far students truly understand the material they present and discuss if they remain so dependent on this technology to express ideas, ask questions and even provide answers.

At this point, the issue is no longer simply whether students use AI, but whether they are still engaging in the thinking process that should lie at the heart of higher education. Education has never been solely about producing assignments that look good.

Behind every piece of writing lies a process of reading, understanding, weighing information, connecting concepts, constructing arguments and finding the right way to communicate ideas. It is this process that develops students’ intellectual abilities.

However, when that entire process is handed over to AI, students may still produce good assignments, but their understanding may not match the quality of the work. Their writing may look academic, the sentences may be well structured and the arguments may seem convincing. Yet when asked to explain the content, defend their arguments or connect them to real-world issues, they may not be able to do so.

I often encounter cases in which a student submits an assignment containing an impressive argument but struggles to explain the meaning of a paragraph they have written themselves. If that happens, what are we actually assessing: students’ ability to understand an issue, or their ability to produce convincing-looking AI output?