In Praise of Learning: Charting the (new-ish) Pathway to Academic Integrity in the Age of AI

What a title. It’s almost competing with “The Persecution and Assassination as Performed by the Inmates of the Asylum of Charenton Under the Direction of the Marquis De Sade” by Peter Weiss for being a ridiculously long title that says exactly what’s on the tin (thank you Ronseal).

I know I’m definitely pushing the limits of SEO here. The title isn’t totally mine, it comes via Tony Frontier, author of ‘AI with Intention’, which you can buy from Amazon here. Alternatively, if you have a Perlego subscription, you can read it there for ‘free’ (you pay for the subscription, so it’s not really free, but you know what I mean!).

The point of this post is to provide another signpost to post-primary school teachers that education is changing (in case you haven’t noticed). AI has already changed education and will continue to do so – in ways that may not be easily predicted or known at the time of writing. That means we have to change how we do, what we do in the classroom in order to be able to prove learning. That won’t be an easy task, but it’s not impossible.

Are you ready? Let’s get to work.

Academic integrity.

That’s it. In many ways, that’s the post. That’s the point and that’s the ballgame (as would be said in American baseball).

It sounds easy, but how do we put this in to our school culture?

Frontier considers integrity as a coin that reads “cheated” on one side and “didn’t cheat” on the other. But integrity is a different coin entirely – integrity can’t really be so black and white in terms of if you cheated, of if you didn’t cheat. He defines cheating and integrity in the following ways:

• Cheating is using a tool or resource to misrepresent one’s knowledge and skills to receive undue credit for a task. The accuser needs to support a claim about cheating with evidence that cheating occurred.
• Integrity describes a commitment to ensure one’s completion of a task accurately represents the knowledge and skills one actually possesses. A claim about integrity is supported by a learner’s ability to transparently document and reflect on the learning process and explain or expand on their results.

There are many tools in the teaching arsenal that can help our pupils’ to develop and maintain their integrity, and to even understand why integrity is so important.

Detecting AI is fraught with difficulty. No commercial platform can/has successfully detected accurate use of AI. I was chatting to my principal recently and told him of an exercise I completed in checking the accuracy of ‘AI checkers.’ I gave AI a prompt for a piece of writing on a topic within educational technology. I took the output, and with no editing whatsoever, put it into an AI checker. The platform returned to me an outcome that the piece of writing was 100% human, despite me knowing it was 100% AI  created.

There are some key takeaways from this:

1. It’s getting harder to differentiate between human written prose and AI written prose in some circumstances.
2. There will be pupils who may be able to ‘slip through’ the net in terms of submitting work that is not their own.
3. There are pupils who will get caught having used AI, submitted and declared work to be their own and as a result, have their grade rescinded in a subject. This will have inevitable consequences for university applications, returning to school for A-Level or in Higher Education, have an impact on degree classification and career prospects.
4. There will be pupils who are accused of using AI, haven’t, and then may have no real way of proving their innocence.

As teachers, it is our job to help pupils mitigate these risks. I appreciate there will be people who will say “don’t use AI” but let’s be real here. If you have a tool, that can help you through an assignment and the potential for getting caught is low because teachers don’t care or aren’t bothered to look to hard, then would you use it? Of course you would. I would too, and anyone who says otherwise is being dishonest.

I appreciate it’s easier for adults currently – we have grown up in a world that required us to learn. The internet wasn’t what it is now when I was at university and I know for a stone-cold fact that had it existed then, the way it does today, my lecturers wouldn’t have had a clue about how it could be weaponised for academic purposes. As a bit of context, I did my dissertation on the religious themes found in The Matrix (first film, and best!) and my study supervisor asked if I had a copy of it on VHS. In 2004. For a further global context, Netflix started their streaming services in 2007. I’m not having a go a the lecturer for not being tech-savvy, just highlighting how easy it could have been to fool some educators.

For pupils though, the risks appear to be higher. Achieving high grades is almost mandatory. I know in my school, in the past I haven’t felt the achievement of pupils getting A grades, just that I met the targets that were expected. Then we get into university. Courses currently cost around £5500 in Northern Ireland per year, around £10,000 in England and then there is pressure of achieving a final classification that will help secure graduate employment. Courses are now weighted throughout the course, instead of a final exam style weighting, meaning that these assignments count towards that end grade. This only considers UK-based pupils. If you’re a pupil in America, then the stakes are even higher – ensuring a high enough GPA, to then gain entry to a top university (and outside the Ivy League universities, I’m not sure how they rank) which will then contribute to your graduate prospects. The risks have never been higher.

I’m explaining this because I know as a teacher, it’s easy to forget what pupils face – and I haven’t even considered issues of part-time employment during university or any issues at home – and only focus on our subject, but the overall picture is much bigger than that.

Returning to the point of this section, detecting AI can be difficult, and if detecting it is difficult, then the potential for a false accusation is greater. And where this is greater, the potential for a pupil to have a defense that is more than “I didn’t use it, I promise” is significantly reduced.

It’s not our job as teachers to police who is using AI and who isn’t, but it is our job to educate pupils in how and when to use AI in ways that are appropriate. When we acknowledge what it can and can’t do, then we can move into the areas in which we can make a difference and educate our pupils in how to make productive use of AI.

As a Computing teacher, there are absolutely ways we can do this using technology. But there are also non-digital means too. We’ll go through a few of them here.

Emphasise the voice of the individual. I’ve said many times I’m a great fan of AI and what it can help us to do in education, both for teachers and learners. There are parts of learning that shouldn’t be determined by AI – our individual voice is one of them. In publishing this site, I have researched, learned, thought, evaluated, written, deleted, written again, deleted, written better and then (finally) published. But the process is what’s  important. It has developed me as a teacher, as a thinker and as a creator. I’m pretty sure if you compare one of my first posts to this one, there should be a difference and a greater level of maturity in how I write. The voice we have is one thing that AI can never take, replicate or replace. Our view on the world, evaluation of facts, combining, assimilating or rejecting world views on the basis of evidence is not easily achieved by AI. When we write, we give a window in the very thought process that makes us human. For that reason, writing is important.

Through the practice and development of writing, we teach our pupils to think, how to formulate a coherent argument, how to breakdown and solve a problem through the process of analysis. By developing this and improving pupils’ ability to think and write, we should inherently reduce the actual need to embark on a path of malpractice – either through plagiarism or the misuse of AI.

I am a great fan of Jordan Peterson. I find a lot of what he says makes sense, even if he is unpopular with some – it’s hard to fault his thought process and logical approach. He is a great teacher who will absolutely make you think! In the video below, he discusses why it’s important to write. It’s only 4 minutes long, but totally worth your time.

This is probably the hardest aspect of all. How can we instill in our pupils a sense of integrity? How can we ensure that when they have an opportunity, or are faced with the question of “should I use AI and not say anything?” that they respond in a way that is academically and morally honest?

Much of how we manage this can come through the culture we develop and set in our classrooms (and by extension, our schools). If you have embedded Rosenshines’ Principles or good questioning into your classroom practice, then there are ways we can test for understanding in small but consistent ways to determine or guide our suspicions when it comes to malpractice. If we take student A, who answers well in class, shows to have a good understanding of the concepts and backs that up through testing, then the production of a good piece of coursework won’t necessarily raise initial suspicions. But student B, who struggles in class, doesn’t perform well in testing and then produces a high piece of coursework, may raise some suspicions. This is not hard and fast, because different people can perform better under different circumstances, but you hopefully get the point – it will point towards an idea that should prompt further questioning.

I can accept that for a sizeable percentage of school pupils that this might not be an issue. I know a lot of my current senior A-Level pupils are ‘anti-AI’. They don’t even like seeing teachers use it (they think it’s ‘lazy’ but given them a month as a teacher and they’ll be using it too). But what about that other group? I don’t really like using this type of generalisation, but morality can be a funny old thing – we can be opposed to something one day and in favour of it the next, particularly when self-interest and pressure comes to the fore. There will be a group of pupils who won’t care about the moral/ethical aspects of this discussion because they see it as helping them to get a grade that they need to get. When that type of rationalisation is used, what do we do? If you don’t even know where to start with that question, then you’re in luck! I have a suggested set of school strategies that can be used in the next section.

4.1 Align expectations with opportunities to learn

When classroom priorities match the learning experiences pupils actually receive, they are better equipped to pursue ambitious and achievable learning goals.

• Share the course overview at the start so pupils clearly understand the key standards, goals, concepts, skills and checkpoints. Having a clear “roadmap” supports planning and is especially important for pupils preparing for external examinations.
• Announce major assignments and assessments early to allow pupils time to organise their workload.
• Ensure formative activities directly support the knowledge and skills required for summative tasks, so pupils have had fair opportunities to learn what will be assessed.
• Use rubrics as a shared learning tool, not just a way to justify grades. If an outcome such as “support claims with relevant details” is required, this should appear in the rubric, be taught explicitly, used during practice, and serve as a focus for feedback.
• Emphasise learning goals rather than flawless performance. Pupils should feel comfortable asking for help and learning from mistakes. A classroom culture built on “getting everything right” encourages task‑completion rather than understanding. Norms for questioning, correcting errors, and reflecting on progress help pupils focus on mastery instead of perfection.

4.2 Affirm pupils at the starting line

Pupils who feel known, valued, and understood by their teachers are more likely to behave with integrity. Positive relationships make pupils think twice before letting their teacher down.

• Use an interest inventory to learn about pupils’ backgrounds, motivations, and learning preferences.
• Show pupils that their interests matter by referring to what they wrote and using this information to connect with them, plan learning activities, or make content more relevant.
• Collect baseline evidence of writing by giving an in‑class task that requires a clear claim and supporting evidence. This helps you understand each pupil’s writing style and skill level.
• Provide acknowledgement and brief feedback on this baseline work (without grading it) so pupils know you recognise their starting point. Positive comments, and supportive guidance when needed, help build confidence and a sense of partnership.

4.3 Create clear boundaries for what resources may and may not be used

Vague rules lead to confusion. In a world where AI tools are widely available, simply saying “don’t use AI” is unhelpful. Clear guidance helps pupils understand expectations and prevents misuse.

• Clearly explain what counts as inappropriate use, such as copying work from any source or tool without proper credit.
• Define acceptable use, including when pupils may use AI tools to brainstorm, summarise, explore ideas, or receive feedback.
• Teach pupils how to use AI responsibly so they see it as a tool for deeper learning—not a shortcut. They need guidance to use AI for meaning‑making, clarifying misunderstandings, and improving their work.
• Maintain consistent expectations for transparency. Pupils should be expected to record and disclose their use of all outside resources, including AI, tutors, peers, and digital tools.

4.4 Acknowledge progress at checkpoints

Checkpoints are meant to support pupils, not penalise them. They help pupils stay on track and make steady progress.

• Break large tasks into smaller steps with staggered deadlines to encourage planning, drafting, revising, and using feedback effectively.
• Use quick in‑class checks for understanding, such as short written responses or exit tickets, to gauge progress and reinforce learning goals.
• Teach pupils how to pace themselves by asking reflective questions such as: “What have you completed so far?”, “What do you need to do next?”, and “What questions remain?”

4.5 Provide consistent expectations for acknowledging sources and outside help

Inconsistent enforcement of academic integrity rules leads to confusion. Pupils need clarity about what counts as assistance and when sources must be acknowledged.

• Teach pupils how to cite sources, including textbooks, websites, and AI tools. Show clear examples and expect consistent use throughout the year.
• Help pupils understand when ideas or contributions from others must be acknowledged—not just how to format citations.
• Require regular source acknowledgment on everyday tasks so pupils can build these skills gradually and confidently before higher‑stakes assessments.

4.6 Seek evidence of transparency and “explainability”

Asking pupils to describe their thinking reveals the authenticity of their work and reinforces integrity as an ongoing expectation.

• Gather transparency evidence with prompts such as:
  • “What strategies did you use to complete this assignment?”
  • “List any people, tools, or resources you used.”
  • “Attach any AI prompts and responses you used.”
• Collect evidence of ‘explainability’ by asking questions like:
  • “Choose a sentence from your work that you are proud of and explain your thinking in more detail.”
  • “What are the three most important ideas from this task, and why?”
  • “Summarise your main points in 50 words.”
• Normalise integrity checks. Evidence should be collected routinely, not only when cheating is suspected. Quick writes or digital prompts during class time work well.
• Ensure you have integrity evidence before assigning marks. If work is significantly stronger than a pupil’s explanations, this may indicate a gap in understanding rather than misconduct. Use this as an opportunity to help the pupil revisit the task and demonstrate true comprehension.

4.7 Frame integrity as the basis of your partnership for teaching and learning

Regardless of your school’s policy, including a clear statement in your course materials (Frontier, 2025) can highlight the essential role integrity plays in the learning partnership between teacher and pupil.

Many teachers have attempted to deter cheating with AI by threatening greater consequences. But increased consequences have a paradoxical effect. The greater the consequences for the cheater, the greater the burden of proof for the accuser. By acknowledging the limits of our strategies that seek to police integrity by catching students cheating, we can turn our attention to a more comprehensive set of strategies that minimise cheating and maximising integrity.

This was a much longer post than many previous posts, but there’s a lot to consider in this one. I know a lot of schools are facing this at the moment (along with FE & HE institutions) and there isn’t a great answer at the moment. I can foresee a lot of careers becoming more practically-minded and graded on the basis of how you do the job under observation, through a demonstration of applied knowledge under ‘pressure’ (this could be a demonstration for example), rather than written essays.

In this way, AI can’t support learner malpractice as evidence of learning will rely on the learner proving their competence – which can only be proven through the practical part of a given task. Presentations with question & answer sections to demonstrate understanding may also become the norm. We know you can use AI to produce an answer, but can you show you understand it? Equally in post-primary education, written exams under timed conditions could also be considered as a solution to AI achieved grades.

Whatever the future will look like, AI will be there and education will adapt to deal with it. There are a lot of the finer details that need to be ‘ironed out’ of how exactly that will look.