Key Takeaways

• AI cheating destroys what school is actually for: building the capacity to face and master difficult things. The grade is the receipt; the capacity is the product.

• Students using AI to escape difficult work are not avoiding the anxiety; they are postponing it to a moment in adulthood when the gap between resume and capability becomes visible.

• The underlying driver is almost always anxiety rather than laziness, particularly fear of not being good enough or of disappointing parents and teachers.

• Treating AI cheating as a discipline problem misses the function the behaviour is serving and produces only compliance until the next opportunity to defect.

• The conversation that works with teens starts with curiosity about what is making the work feel impossible, not with confrontation about whether the rule was broken.

AI cheating destroys what school is actually for: building the capacity to face and master difficult things. Students who use AI to do their work get the grade and lose the only thing that actually mattered, which is the developmental experience of struggling with something hard and discovering they can handle it. The grade is the receipt. The capacity is the product. When you skip the capacity, you have bought nothing, even if the receipt looks impressive, and the real cost of that trade does not arrive until years later when the consequences are much harder to undo.

I'm Dr Rick Smith, PsyD | EdD, a clinical psychologist in Hong Kong who spent nearly two decades in classrooms before moving into psychology, working with international school families and high-performing teens on anxiety, executive functioning, and the kind of avoidance that has now found its most efficient delivery mechanism in artificial intelligence. The number of teens I see whose anxiety is being fed and protected by AI shortcuts has grown steadily over the past two years, and the pattern is consistent enough that I now bring it up early in family conversations whether or not the parents have raised it.

What is AI cheating actually costing students?

AI cheating costs students the only thing school was supposed to give them: the experience of doing something difficult, finding it almost too hard, and discovering they could do it anyway. That experience is what builds the durable confidence that holds up under pressure. A grade is a receipt for that experience. Skip the experience and the receipt is empty, even if the transcript looks identical to the student next to you. Every essay written by AI rather than by the student is a missed repetition of the muscle that essay was supposed to be building. Every problem set solved by AI is a missed encounter with the productive struggle that makes the next problem set slightly easier. Across hundreds of these missed reps over years of schooling, the student arrives at university or first job with the appearance of preparation and almost none of the underlying capability. They have the resume of someone who can do hard things and the nervous system of someone who has never had to. That gap closes the first time they hit a problem with no prompt available, and the closing is brutal. Real anxiety in adulthood is rarely about grades. It is about discovering, often in a high-stakes setting, that you do not have the skills the room assumes you have. Students using AI to escape that anxiety in school are not avoiding it; they are postponing it to a moment when it will be much more expensive to face.

Why are students reaching for AI instead of doing the work?

Students reach for AI for the same reason adults reach for any avoidance behaviour: it removes immediate discomfort. Sitting down to a difficult assignment produces a small wave of unpleasant feeling, including the prospect of struggle, the possibility of failure, the boredom of slow progress, the comparison to peers who seem to find this easier. AI offers an immediate way out of that feeling. Type the prompt, get the output, submit it, feel the relief of having dispatched the problem. The nervous system registers the relief and learns the pattern. The next time a similar feeling appears, the pull toward AI is slightly stronger, and the willingness to sit with the discomfort is slightly weaker. Over weeks and months, this is how an occasional shortcut becomes a default coping strategy. The underlying driver is almost always anxiety rather than laziness. The students I see who have fallen into chronic AI use almost universally describe a felt sense of being behind, of not being smart enough, of not being able to produce work as good as their peers, of dreading the disappointment of teachers or parents if they submit something imperfect. AI removes that fear in the moment, which is exactly why it is so effective as an avoidance tool and exactly why treating it as a discipline problem misses the underlying issue. For students whose anxiety has begun shaping how they engage with academic work, the AI use is the symptom; the anxiety is the condition.

Why is treating AI cheating as a discipline problem the wrong response?

Treating AI cheating as a discipline problem is the wrong response because discipline frameworks address the behaviour without addressing the function the behaviour is serving. When a school catches a student using AI inappropriately and responds with academic penalty, the immediate behaviour is interrupted, but the underlying anxiety that produced the behaviour is left in place. The student now has the original anxiety plus the shame of being caught, and the next time they face a difficult assignment they have an even stronger reason to find a more sophisticated way to use AI without detection. The discipline approach also misses the more important conversation. The question is not whether the rule was broken; the question is what was the student trying to escape, and what would it take to help them face that thing without escape next time. Schools that handle this well combine clear academic integrity standards with structured support for the underlying anxiety and skills gaps that drive most AI cheating. They do not abandon consequences; they pair consequences with the kind of work that actually changes the pattern. For families, the move is the same. The question worth holding is not how do I stop my teen from using AI inappropriately. The question is what is the AI doing for them that they cannot yet do for themselves, and how do we build the capacity that would make the AI unnecessary.

What can parents actually do if they suspect their teen is using AI to avoid difficulty?

Parents can do four things in combination, and the order matters. The first is to have an honest, non-punitive conversation with the teen about what is happening, focused on understanding rather than on catching them out. The opening line that works in my office sounds something like: I am noticing some things about how you are managing schoolwork and I want to understand what is going on for you, not to punish you, because the punishment is not going to fix what is actually happening. The defensive layer usually drops within a few minutes once the teen understands they are not being trapped. The second move is to ask what is making the work feel impossible. Anxiety about not being good enough is the most common answer. Perfectionism, fear of disappointing parents or teachers, and overwhelm from accumulated avoidance are the others. Naming the actual driver is most of the work, because once it is named, it becomes addressable rather than mysterious. The third move is to build small repetitions of the experience the teen has been avoiding, starting at a level low enough to succeed at. This often means picking one assignment to do entirely without AI, of any length or complexity, just to give the nervous system the experience of completing something hard. The fourth move is to address the underlying condition if the pattern is severe or longstanding, which often means working with a clinician on the anxiety, perfectionism, or avoidance and procrastination pattern that is driving the AI use. The work is usually shorter than parents expect, three to six months for most teens, and the change is visible in the teen's relationship with effort long before it is visible in the grades.

When does AI avoidance become a clinical concern rather than a behaviour problem?

AI avoidance becomes a clinical concern when it is part of a larger pattern of avoidance rather than an isolated shortcut, when the teen's anxiety has begun shaping not just academic work but the rest of their life, when previous attempts to address the behaviour have not held, or when the teen has begun to express the kind of demoralisation that suggests they have stopped believing they can do the work themselves. Hold the line on family-led change when the AI use is occasional rather than chronic, when the teen can still engage with difficult tasks without escape when asked, when academic performance has not yet declined significantly, and when the teen still believes they are capable. Bring in support when the pattern has become pervasive, when sleep, mood, or relationships have started to slip alongside the academic picture, when the teen has begun to avoid school more broadly, or when the anxiety underneath the AI use is producing physical symptoms or significant distress. Most of the teens I see for this pattern respond well to a structured course of work that addresses both the avoidance behaviour and the underlying anxiety. The behavioural change is usually faster than the anxiety change; the anxiety change is what makes the behavioural change durable. A first consultation typically clarifies, in about thirty minutes, which layer needs work first, and from there the path is much clearer than it looks from inside the family.

The honest summary is that AI cheating is not really about cheating. It is about a generation of high-pressure students discovering the most efficient avoidance tool ever made and using it to escape difficult work, while losing the developmental experience that the difficulty was supposed to produce. If your teen has fallen into this pattern, the work is genuinely tractable, but it has to start with what is driving the avoidance rather than with the avoidance itself.

Frequently Asked Questions

What is AI cheating actually costing students?

It costs them the only thing school was supposed to give them: the experience of doing something difficult and discovering they could handle it. That experience is what builds durable confidence. The grade is the receipt; the capacity is the product. Students who skip the capacity arrive at university or first job with the resume of someone who can do hard things and the nervous system of someone who has never had to, and the gap closes brutally the first time they hit a problem with no prompt available.

Why are teens using AI to do their schoolwork?

Almost always because it removes immediate discomfort. Difficult assignments produce a small wave of unpleasant feeling, including the prospect of struggle and the possibility of failure. AI offers an immediate way out. The relief reinforces the behaviour, and over weeks and months an occasional shortcut becomes a default coping strategy. The underlying driver is almost always anxiety rather than laziness, particularly fear of not being good enough or of disappointing parents and teachers.

Is treating AI cheating as a discipline problem effective?

Discipline addresses the behaviour without addressing the function the behaviour is serving. When schools catch and penalise AI use, the immediate behaviour is interrupted but the underlying anxiety is left in place, and the student now has both the original anxiety and the shame of being caught. Schools that handle this well combine clear academic integrity standards with structured support for the anxiety and skills gaps that drive most AI cheating. The two approaches work better in combination than either alone.

What should parents do if they suspect their teen is using AI to avoid difficulty?

Four moves in combination: have an honest non-punitive conversation focused on understanding rather than catching them out, ask what is making the work feel impossible and listen for anxiety, build small repetitions of the experience the teen has been avoiding starting at a level low enough to succeed at, and address the underlying anxiety or avoidance pattern with a clinician if the issue is severe or longstanding. Naming the actual driver is most of the work, because once it is named, it becomes addressable.

How can I tell the difference between occasional AI use and a real problem?

Occasional use is contained, situational, and does not affect the teen's relationship with difficult work in general. A real problem is pervasive, has become a default response to discomfort, has begun affecting sleep, mood, or other areas of life, or coincides with declining confidence in the teen's own capability. The behaviour itself is less diagnostic than its function: is the AI removing the occasional difficulty, or has it become the way the teen avoids difficulty altogether.

When should we bring in a clinician about AI avoidance?

When the pattern is pervasive rather than isolated, when sleep or mood or relationships have started to slip alongside the academic picture, when the teen has begun avoiding school more broadly, or when the anxiety underneath the AI use is producing physical symptoms or significant distress. Most teens with this pattern respond well to a structured course of work within three to six months. The behavioural change is usually faster than the anxiety change; the anxiety change is what makes the behavioural change durable.

Author bio

I'm Dr. Rick Smith, a clinical psychologist in Hong Kong working with high-performing teens and adults on ADHD, anxiety, OCD, addiction, and executive functioning. My work draws on Acceptance and Commitment Therapy, Cognitive Behavioural Therapy, and Exposure and Response Prevention, applied to international school families and the expatriate community.

Before psychology, I spent nearly two decades in classrooms supporting students with learning differences. I'm the author of STOP Reading (4.8 stars on Amazon) and deliver workshops for schools and organisations across the region. More at rick-smith.com.