Most articles about AI in education say the same thing: use it to mark essays, write reports, or build entire schemes of work. And while that’s all fine, I think English teachers deserve better. They need more creativity, more relevance, and more options that actually work in the classroom.
I’ve been playing around with AI for a while now. I’m not interested in replacing what teachers do, and I don’t think students should be using it to write full essays either. But I do think there’s a place for AI in education, both for teachers and students, if we use it well and I honestly believe that in itself is a skill.
I’m curious about it. I’m excited by it. And I’ve found it can be a surprisingly powerful tool, especially in creative writing and literature lessons. Here are 7 tried-and-tested ways I’ve used AI in the classroom.
Create Original Picture Prompts for Creative Writing
I’ve always believed a picture is worth a thousand words. Before AI, I’d spend hours scouring the internet for the right image. I would be looking for something that felt atmospheric, specific, and not like the same five stock photos I’d seen in every English classroom. It was especially tough for genres like dystopia, magical realism, or sci-fi. The ideas were there… but the visuals? Not so much.
Now, I come up with the concept, and AI does the hard work for me.
Foggy train stations, glowing orbs, empty houses at dusk, you can generate visuals that feel like story seeds. Images that stir up atmosphere, questions, or that unsettling what’s going on here? feeling.
And because you’re not limited to what already exists online, you can tailor your prompts to the group in front of you. One class might respond best to something eerie and surreal. Another might prefer something warm and romantic. With AI, you can create both in minutes.
It’s genuinely saved me hours of planning time and has helped students feel instantly inspired to write.
Here’s an example of a prompt I used recently:
"Create a mysterious photo prompt showing a lone figure in a foggy forest holding a lantern. Make it look cinematic and surreal."
Even though all the images below came from the same prompt, each one can inspire a completely different story, depending on what details students latch onto. That’s the beauty of open-ended prompts: the creative possibilities multiply.
Want to write your own AI prompts? Here are a few tips to get started:
◆ Focus on mood or genre (e.g. eerie, nostalgic, dreamlike)
◆ Mention specific details like lighting, time of day, or objects
◆ Include just enough mystery to raise a question, without giving away the whole story
◆ Use sensory language if you're stuck, such as fog, flickering light, cracked pavement, etc.
Once you’ve got the hang of it, it’s honestly addictive.
If you’d rather skip the trial and error, I’ve created a full library of image-based writing prompts with story starters and plot sparks - designed specifically for busy teachers. If you want a month's worth of those prompts for free, you can explore those here.
Produce Model Paragraphs for Students to Mark
I used to write model paragraphs all the time, both live in lessons so students could see my thought process, and in my planning time to prep for key moments in a unit. I still do that when it really counts. But now, I often start with AI.
I give it the structure we use (so for us that’s P.E.T.A.L.), feed in the mark scheme we’re working to, and ask for three versions: a weaker one, a mid-range one, and a high-level response. It saves time, and a lot of it.
I get my students to use the mark scheme to estimate a grade, and they genuinely seem engaged when comparing them. They’re more honest (and often more brutal) when they’re not marking each other’s work. You get proper critique:
“She’s just retelling.”
“That one doesn’t even explain the quote.”
“They’ve used a connective, but it doesn’t actually go on to analyse anything else”
Suddenly, they’re not just learning how to write; instead, they’re learning how to read writing like a marker would. That’s where the confidence starts to grow, and the understanding of what they need to do to reach the exam standard they are aiming for.
Here’s one of the prompts I’ve used recently and the paragraph AI produced for me:
“Write a high-level PETAL paragraph answering the question: ‘How does Orwell use the character of Napoleon to explore ideas about power and control in Animal Farm?’ The paragraph should be of GCSE standard, use embedded evidence, and include precise analysis of both language and theme.
Use the mark scheme provided to guide the level of depth and accuracy expected.”
This was a solid example I could quickly tweak. We annotated it together in class, looking at:
vocabulary choices
embedded quotes
analytical depth
what made it feel high-level (rather than just long)
Want to write your own? Here are a few tips for prompting AI effectively:
◆ Always give the structure, e.g. “Use PETAL” or “Follow PEEL format”
◆ Be specific about the question and text because clarity matters
◆ Tell it the grade band and what level of analysis to aim for
◆ Add markers like: “use embedded evidence,” “include analysis of language,” or “link to theme”
◆ Start with one version (high or mid), then ask it to simplify or stretch that same idea
Once you’ve got a few strong examples, you’ll start building a library of model responses tailored to your class, without doing all the heavy lifting yourself.
Roleplay as Characters or Authors
I’ve always used hot seating in literature. It’s one of those low-prep, high-impact activities that gets students thinking about character motivation, voice, and they tend to really enjoy it. Normally, I’ll model a response, and then students take turns either as the character or as the questioner.
Using AI is just an evolution of that. You can simulate responses as Macbeth, Offred, Ray Bradbury - literally anybody. And let students interrogate the character or author in a more structured way. It adds depth, especially if you’re modelling how to dig beneath the surface. I still do the classic hot seat setup, but using AI as the character gives you more to bounce off and more angles to explore, which leads to stronger discussion.
So far, we’ve done this as a whole class, but some students have also played around with it at home, and the responses it gives often become great discussion starters, or makes them think about the text we are looking at from a different angle
It’s especially useful in literature lessons where you want students to think critically about:
Character voice and decisions
Authorial intent
Power, morality, and unreliable narration
Themes and context, in their own words
Below are some of the answers AI has generated in character for my classes, including Macbeth, Offred from The Handmaid’s Tale, and as Ray Bradbury
Use AI to Generate Practice Paragraphs with Common Mistakes
This is one of the most practical ways I’ve used AI in planning, creating sample paragraphs that include typical errors students make when first learning how to analyse texts.
They’re not meant to be “bad” paragraphs, instead just ones that almost work. Maybe the quote is dropped in rather than embedded. Maybe the explanation repeats the point without going deeper. Maybe the link to the theme is missing entirely. It’s not about laughing at mistakes, it’s about learning to recognise them and how to correct them.
When students practise identifying what’s missing, they start to see those patterns in their own writing, too. It helps demystify what teachers mean when we say “go deeper” or “analyse more precisely.”
You can use this kind of task as a starter, a revision warm-up, or get them in pairs to assess the paragraph together. Sometimes we annotate together as a class, and sometimes I ask students to rework the paragraph, either in full or just by rewriting a sentence at a time.
I used this AI prompt to generate the paragraph below:
"Write a PETAL paragraph answering the question: ‘How is power shown in Macbeth?’ Include a basic point, a dropped-in quote, and a surface-level explanation. It should sound like a student who’s starting to grasp the structure but hasn’t quite mastered analysis."
Tips for writing your own practice paragraph prompt:
◆ Be precise in your wording by using “emerging writer” or “developing paragraph” to set the tone
◆ Focus on just one or two areas for improvement per example
◆ Pair it with a stronger version so students can compare and reflect
◆ Encourage annotations before rewriting, so for example what’s working? What could be clearer?
◆ Use language that builds confidence - this isn’t “wrong,” it’s “in progress”
This is a brilliant way to build assessment literacy without making students feel self-conscious. It also opens up space for less confident students to engage, especially if they’re more comfortable analysing writing than producing it straight away.
Turn Student Ideas into Story Seeds
One of the simplest ways to bring AI into creative writing is by using it as a story seed generator, built from your students’ own ideas.
I’ll often ask the class or struggling students to throw out three elements:
a setting
a mood
and an object
From that, I feed the details into AI and ask it to generate a writing. The result is something cohesive, imaginative, and instantly usable, even when the original ingredients are completely random. And I can also build in additional creative support by asking it to also generate a picture prompt
The beauty of it is that it gives students a creative starting point that still feels like theirs. I’ve had brilliant openers generated from the most random of combos too, which I think shows students that literally any of their ideas can become a piece of creative writing.
It’s ideal for building engagement with reluctant writers or for injecting new life into warm-up tasks and one-off lessons. You can build full-class prompts together or challenge students to work in small groups to develop their own. It’s also an engaging and fun way to help support lower-ability students and make them feel like they still have ownership over the creative process.
I recently finished a Fantasy Writing unit with my little year 7s, and one of my groups helped me create the following creative writing prompts (that all link together) from their ideas.
Tips for building great story seed prompts:
◆ Ask for student input first as it builds ownership before the writing even starts
◆ Choose varied categories, so setting + object + mood = strong mix
◆ Keep the structure simple, story starter or prompt, not a full story
◆ Encourage editing or reworking, as students can write the AI’s idea from a different POV or in a different tense
◆ Use the same process to generate multiple versions, as it is perfect for creative choice
This works brilliantly in creative writing units, after assessments, or as a quick, low-stakes way to build confidence. It turns your classroom into a collaborative writing room, and that energy tends to stick.
Differentiate Tasks On Demand
Every class is a mix. You’ve got the ones who finish in five minutes and ask for more, and the ones who need the question broken down three different ways before they feel safe starting. Differentiation is essential, but it’s also unfortunately time-consuming.
AI makes that a little easier.
I’ve started using it to adapt tasks based on the needs of the group, either in the planning stage or mid-lesson, if something clearly isn’t landing. You can feed it a core task or question and ask for:
a scaffolded version for emerging writers
a challenge task for confident students
or a slightly reworded version for students with EAL or additional needs
You still have control, you're not handing over your lesson to a chatbot (plus I am far too much of a control freak to do that!), but it gives you a head start, especially if you teach mixed-ability groups or set classes across a key stage.
Here’s a sample prompt I used for Year 9 earlier in the year:
"Create three versions of the same literature question on The Tell-Tale Heart: one scaffolded for struggling students, one standard GCSE-level, and one extended GCSE response for high achievers. Focus on how tension is built in the story."
Tips for creating differentiated tasks using AI:
◆ Start with your original question, then ask for variations by ability
◆ Be clear about what you want to adjust such as vocabulary, sentence starters, response length
◆ Use Bloom’s Taxonomy language (describe, explain, evaluate, etc.) to shape challenge
◆ Check everything - it’s a tool, not a teacher
◆ Save the results and reuse them as it’s great for revision or carousel activities later
Once you’ve got a few examples, you’ll start seeing how easily one core task can be stretched or simplified to meet a whole range of needs, and you don’t need three hours of prep to do it.
Personalised Book Recommendations
Sometimes the biggest challenge isn’t teaching analysis or essay structure, it’s getting a student to actually want to read.
I’ve started using AI to help with this. When a student says something like, “I liked The Hunger Games but I hate reading,” or “I want something like Holes but more modern,” I’ll feed that into AI and ask for a few tailored suggestions. The results often give me options I wouldn’t have thought of on the spot, especially for reluctant readers or students with very specific tastes.
It’s not a replacement for a well-stocked library or teacher knowledge, but it’s a great starting point when a student needs a nudge. You can use it live in class, during library lessons, or even with parents at parent-teacher evenings when they ask, “What should they read next?”
Tips for using AI to support reading for pleasure:
◆ Include student age, reading ability, and interests for best results
◆ Reference books, films, or games they already enjoy
◆ Ask for content warnings or maturity level if needed
◆ Use it to build a class reading wall or themed book lists
◆ Encourage students to try the prompt themselves, so it builds ownership and curiosity
This isn’t about turning AI into a librarian, it’s about using the tools we’ve got to help students find books that they actually connect with. Once they’ve read one they enjoy, they’re far more open to the next.
And on a side note, it also works just as well for TV and movie recs when you need some downtime!
Final Thoughts
AI isn’t a miracle fix, and it’s definitely not here to replace teachers. But used intentionally, it can be a brilliant tool to save time, support differentiation, and keep creativity alive in the classroom. Whether you’re planning a lesson, building confidence through model responses, or just trying to get a student excited about reading again, I think it’s worth exploring.
You don’t have to use all seven ideas at once. Start with the one that feels manageable, and see what works for your teaching style.
