Blog / Tech, New & Old /
AI in schools: help or shortcut?
My view: AI in schools is neither a miracle nor a cheating machine. It’s a tool that makes good learning better and lazy learning easier, and which one you get depends almost entirely on how schools set it up.
The honest answer to “help or shortcut?” is “both, and the system decides.”
Where it helps
A patient tutor
A student stuck on a maths step at 10pm has nobody to ask. An AI tool can explain it again, in a different way, as many times as needed, without anyone feeling judged. For a shy student, or one whose family can’t afford a tutor, that’s a real gain.
Language and accessibility
AI can simplify a hard text, translate, read aloud, or help students whose first language isn’t the language of instruction. It can help students with dyslexia or other learning differences by changing how information is presented.
Practice and feedback
Generating practice questions, quizzing a student on a topic and giving instant feedback on a draft are things it does well. Teachers can’t give every student feedback on every attempt; a tool can fill part of that gap.
Saving teachers time
Drafting lesson plans, worksheets and differentiated materials can take hours. If AI reduces that, teachers have more time for the part only humans do: noticing, encouraging and teaching.
Where it becomes a shortcut
Doing the thinking for you
The point of an essay isn’t the essay. It’s the thinking required to write it. If a student pastes the question into a chatbot and hands in the result, they get the grade and skip the learning. And they may not notice what they missed until an exam, or a job, shows it.
Confident mistakes
AI tools can state wrong things fluently, including made-up quotes and sources. A student who trusts without checking learns wrong facts with great confidence. I wrote about this in a coding context in AI hallucinations in code, but the problem is the same for history essays.
Weakening basic skills
If you never practise writing a paragraph, summarising a text or working a sum, those skills fade. There’s a legitimate worry here, even if it’s sometimes overstated.
Unequal access
Students with better devices, paid versions and knowledgeable parents can get far more from these tools than those without. Uneven access can widen gaps instead of closing them.
Why detection isn’t the answer
Many schools reach first for AI detectors. The trouble: these tools are unreliable. They produce false accusations and can be fooled. In 2023, OpenAI withdrew its own AI text classifier because of its low accuracy. Detectors have also been criticised for flagging writing by non-native English speakers more often. Treating a detector’s score as proof is risky and unfair.
A policy built on catching students is a fragile one. It’s better to change the work itself.
What works better
Assess the process, not just the product
- Ask for outlines, drafts and notes along the way.
- Have students explain their work aloud in a short conversation. A student who wrote their own essay can discuss it.
- Include some in-class writing and exams where AI isn’t available.
Teach how to use it
Students will use these tools anyway. Better they learn to do so well:
- Ask the tool to explain, not to answer.
- Check claims against reliable sources.
- Compare its answer with their own, and say what differs.
- Keep a note of how they used it, and be open about it.
Set clear rules
“Don’t use AI” is vague and often ignored. Clear guidance for each assignment works better: where it’s allowed, where it isn’t and how to say you used it.
Design tasks AI does poorly
Work based on local examples, class discussions, personal observation, or recent events tends to be harder to outsource.
The strongest counter-argument
“Calculators didn’t ruin maths, and spell-check didn’t ruin writing. AI is just the next tool, and banning it is pointless.”
It’s a strong point, and I partly agree. Schools did adapt to calculators, search engines and Wikipedia. Banning AI is neither realistic nor wise, because students will meet it at work.
But there’s a difference worth taking seriously. A calculator does a mechanical step; AI can do the whole task, including the reasoning and the writing. The right question is what students still need to be able to do themselves. We let calculators in once students knew arithmetic. The same logic suggests building core skills first, then adding AI where it extends them. It’s why I still think learning to code in the age of AI is worth it: the tool is much better in the hands of someone who understands the basics.
What parents can do
- Ask, don’t accuse. “Show me how you used it” opens a conversation.
- Encourage explaining. If your child can teach you what the essay says, they learned it.
- Check the claims together. Verifying a fact is a skill in itself.
- Look at the school’s policy and ask what’s allowed.
The web admin angle
I work with websites and information, and the same pattern shows up there. AI helps most when you already know enough to judge its output, and does the most damage when it replaces that knowledge. Schools are trying to produce people who can judge. That’s the goal worth protecting.
Key takeaways
- AI helps as tutor, language aid, practice partner and time-saver for teachers.
- It becomes a shortcut when it replaces thinking, and it can state falsehoods confidently.
- Detection tools are unreliable; assess process and understanding instead.
- Teach core skills first, then use AI to extend them, with clear rules.
FAQ
Should schools ban AI?
A blanket ban is hard to enforce and ignores how students will work later. Clear, task-specific rules usually work better.
How can a teacher tell if a student used AI?
Not reliably from the text alone. Conversations about the work, drafts and in-class tasks are stronger evidence.
Is it cheating to use AI for ideas?
It depends on the school’s rules. Whatever the rule, being open about it is the safest course.
Related reading
If you were setting the rule for one assignment, where would you draw the line between help and shortcut?