Artificial intelligence (AI) has not made homework irrelevant. It has made us ask a question we should perhaps have been asking all along: what is homework for?
A pupil is given a homework task on Monday. By Monday evening, an AI tool has produced a complete answer. The pupil copies it, makes a few changes and submits it. The teacher receives a polished piece of work. The pupil receives a mark or feedback. Neither can be entirely confident what has been learned.
This scenario has prompted understandable concern about cheating and academic integrity.
But it also exposes a deeper question. If a homework task can be completed without the pupil thinking, what educational purpose is it serving? AI did not create every weakness in homework, but it has made some of them impossible to ignore.
Homework needs a clearer purpose
Homework can serve several valuable functions. It can help pupils retrieve and consolidate knowledge, practise skills, prepare for future learning, read more widely or develop independence. However, not every task does these things equally well.
A lengthy written response may look substantial while revealing little about what a pupil understands. A short retrieval exercise may have greater learning value. Reading a chapter in preparation for discussion may be more useful than producing a generic summary that AI can generate instantly. The question should not be how much work pupils complete outside school. It should be what learning the work is intended to secure.
Stop making detection the whole strategy
Schools understandably want to protect academic integrity. Pupils should not present AI-generated work as their own. But a homework policy built entirely around detecting AI use is unlikely to provide a lasting answer. AI-generated content is becoming increasingly difficult to distinguish reliably from human writing. More importantly, a focus on detection can distract from the design of the task itself.
A better approach is to be explicit about when AI is permitted, when it must be acknowledged and when work must be completed independently. Pupils need to understand why those distinctions matter.
Homework that remains worth doing
AI creates an opportunity to reconsider homework design. Retrieval practice, independent reading and carefully selected questions can help pupils strengthen knowledge. Preparation tasks can give them something meaningful to bring to the next lesson. Other activities might explicitly incorporate AI: pupils could critique an explanation, identify errors, compare interpretations or explain why they rejected a suggested answer. In each case, the task should make the pupil's thinking central.
A useful test for teachers is: could a pupil complete this task successfully without developing the knowledge or skill I want them to learn? If the answer is yes, the task may need redesigning.
Equity matters here too
Homework has always been affected by differences in home circumstances. Some pupils have quiet spaces, reliable devices and adults who can help. Others do not. AI may provide valuable support for pupils who lack that assistance. But it could also create new inequalities between those who know how to use it productively and those who rely on it to complete tasks without understanding. Schools therefore need to teach responsible use explicitly and avoid assuming that all pupils have the same access or support.
Not the end; this is a new beginning
At AiEd, we do not believe AI means abandoning homework. We believe it presents an opportunity to make homework more purposeful, equitable and educationally valuable.
Teachers should be able to explain why a task matters, what pupils are expected to do independently and how any AI assistance contributes to learning. Pupils should understand that producing an answer is not the same as developing understanding.
Perhaps the most useful question for schools is no longer, “How do we stop pupils using AI for homework?” It is this: “What homework is worth doing in an AI-rich world?” The answer could lead to better homework than we had before.