AI and English Literature: using it without letting it think for you

Artificial intelligence has changed the educational landscape, and pretending students will not use it is neither realistic nor useful. The more important question — for parents, for students, and for anyone teaching English Literature right now — is how it gets used.

What AI is actually good at, and what it isn't

AI can produce a polished-looking Literature response remarkably quickly. It can generate a plausible-sounding paragraph on almost any set text, structured correctly, using the right terminology, arriving at a confident conclusion.

None of that means the response is perceptive, accurate, or genuinely well supported. A fluent paragraph is not the same as a convincing argument. AI-generated analysis frequently asserts a reading of a text without the close, specific evidence that would actually justify it — it sounds right without necessarily being right.

This matters more in English Literature than in some other subjects, because the whole discipline is built on the gap between a plausible-sounding claim and a genuinely well-evidenced one. Learning to tell the difference is not a side skill — it is close to the entire point.

Using AI as something to interrogate, not something to copy

Where AI becomes genuinely useful is when a student treats its output as material to test, rather than an answer to accept.

A student might take an AI-generated interpretation of a poem and ask: does the textual evidence actually support this claim, or does it just sound confident? They might identify an unsupported assertion, spot where the analysis stays at the surface rather than engaging with how a specific word or structural choice creates meaning, or notice where a seemingly sophisticated paragraph doesn't actually answer the question being asked.

Used this way, AI becomes another object of critical scrutiny — the same kind of scrutiny a student should already be applying to a study guide, a study group's shared notes, or an examiner's report. The exercise is not learning how to get AI to do the work. It is developing the judgement to recognise when work — anyone's, including AI's — is weak.

Why this approach matters more than a blanket rule

Some schools and tutors take a strict no-AI-at-all approach. I understand the instinct, but I don't think it reflects how students will actually encounter this technology, now or later. A student who has never learned to evaluate AI output critically is not protected from it — they are simply unprepared for the moment they inevitably use it anyway, whether at school, in further study, or beyond.

The more durable skill is judgement: knowing what a strong argument actually requires, recognising when a claim isn't backed by evidence, and being confident enough in one's own reading of a text to disagree with something that merely sounds authoritative.

That judgement is built the same way any analytical skill is built — through close reading, through developing independent interpretation, and through the kind of sustained, individual attention that structured tuition is designed to provide. AI does not replace that process. If anything, it makes the process more necessary, not less.

What doesn't change: the exam itself

The approach described above — using AI to interrogate, question, and test analysis rather than to generate it — has no place in coursework, assessed writing, or examinations themselves. GCSE and IGCSE examinations are sat without any access to AI, and coursework or non-exam assessment components carry a clear requirement that the work submitted is a student's own. Passing off AI-generated writing as a student's own work in an assessed piece isn't a grey area — it's the kind of thing exam boards actively check for, and being caught has real consequences, from a piece of work being disallowed to the qualification itself being affected.

But leaving aside the risk of being caught, there's a more basic problem with doing this: it doesn't actually work. An exam is precisely the moment a student has to read an unfamiliar passage and think, unaided, in real time. A student who has leaned on AI to write their practice essays rather than to interrogate their own thinking arrives at that moment having practised the wrong skill. The exam doesn't test whether a good essay can be produced somehow — it tests whether this student can produce one. Using AI to skip that practice isn't a shortcut. It's a way of quietly failing to prepare, while feeling as though preparation has happened.

The student's own judgement remains central

Whatever role AI plays in a student's wider study habits, the aim in tuition stays the same: developing a student who can read attentively, think independently, and construct an argument they could genuinely defend — regardless of where a first draft or a stray idea came from.

AI should support critical thought, not replace it.