Do Not Let AI End Take-Home Learning

A learner spends a week investigating a difficult question, consulting sources, discussing the problem with colleagues and revising a thoughtful response at home. Another learner submits an equally polished response generated in minutes by artificial intelligence. The two products may look alike. The learning behind them is not.

This uncertainty has led some education leaders to an understandable conclusion: bring assessment back into controlled rooms. If unsupervised work can no longer establish who knows what, perhaps take-home assessment should disappear.

That response may secure one kind of evidence while damaging other educational purposes. Take-home work supports reading, research, reflection, authentic projects, flexible participation and the development of complex products over time. It is especially important to distance learners, working adults, caregivers, learners with disabilities and students who cannot repeatedly travel to a campus.

Artificial intelligence has not made take-home learning obsolete. It has made one submitted product insufficient proof of competence. The answer is a stronger assessment architecture, not the abandonment of learning beyond the examination room.

On 10 August 2026, reporting from Australia said the New South Wales deputy premier and education minister proposed ending unsupervised take-home assessment as a response to artificial-intelligence misuse. The same day's policy discussion included South Australia's announcement of a royal commission intended to examine AI's opportunities and risks across education, public services, health, industry and culture.

These are…