AI Chatbots and Students: What Schools Need to Know
Your student's most trusted source of support might not be a person.
More young people are turning to AI chatbots when they are struggling, asking them for advice, support and simply a place to talk.
This is an issue schools need to be talking about now.
AI chatbots can respond instantly, remember the context of a conversation and communicate in a way that feels supportive and personalised. For a young person who is upset, lonely or unsure who to talk to, that can make an AI chatbot feel surprisingly easy to trust.
But an AI chatbot is not a counsellor, psychologist or trusted adult.
It can sound caring without actually understanding what a student is experiencing. It can provide an answer that feels reassuring without that answer being accurate, appropriate or safe.
That distinction needs to become part of AI literacy for students.
Why are students turning to AI chatbots for support?
There are obvious reasons why a student might choose to talk to an AI chatbot.
It is available 24/7. There is no waiting for an appointment. A student does not have to worry about being judged, embarrassed or having to explain everything to another person.
And the chatbot responds immediately.
Research published in 2026 found that almost one in five US adolescents and young adults aged 12 to 21 reported using AI chatbots for mental health advice in 2025. Most of those who used chatbots for this purpose had not told anyone else they were doing so.
This matters for schools because students may already be using AI for emotional support without teachers, parents or other trusted adults knowing about it.
That does not mean every student who talks to an AI chatbot is in danger.
It does mean we need to understand what can happen when a student starts treating an AI chatbot like a trusted source of personal advice.
AI chatbots can sound supportive without being able to provide real support
One of the biggest challenges with AI chatbots and students is that the conversation can feel much more human than the technology actually is.
A chatbot can say something that sounds empathetic. It can remember what a student said earlier in the conversation. It can ask follow-up questions and tailor its responses.
That can create a strong impression that the chatbot understands the person sitting on the other side of the screen.
But sounding supportive is not the same as providing appropriate support.
The American Psychological Association has warned that general-purpose generative AI chatbots do not have the qualifications and capabilities required to provide mental health care in most situations. It also highlights particular concerns for young people, including the possibility that adolescents may place too much trust in AI or struggle to distinguish simulated empathy from genuine human understanding.
This is an important AI literacy lesson.
Students need to understand that a confident, friendly or reassuring answer is not automatically a trustworthy answer.
AI chatbots and students: understanding the limitations
When I talk about AI literacy with students, I do not think the conversation can stop at how to write a better prompt or how to use AI for schoolwork.
Students also need to understand when not to rely on AI.
That includes situations where they are asking for advice about their mental health, relationships, serious personal problems or anything where the consequences of getting the answer wrong could be significant.
There is a big difference between asking an AI chatbot to explain a concept for homework and asking it what to do when you are struggling emotionally.
The first is a learning task.
The second requires context, judgement, human understanding and, where appropriate, professional support.
AI does not have the same responsibility or understanding as a trusted adult or qualified professional.
The APA's 2026 guidance also recommends that people do not rely on AI as their only source of mental health support and that information from AI should be checked with an appropriate health professional.
This is an AI online safety issue too
We often talk about online safety in terms of strangers, scams, inappropriate content, privacy or social media.
AI needs to be part of that conversation now.
A student may not think of an AI chatbot as an online safety risk because there is no obvious stranger on the other side of the conversation.
But students still need to think critically about:
Who or what is giving me this advice?
How do I know the information is accurate?
What does the AI actually know about my situation?
Am I treating the chatbot like a person?
Should I be talking to someone I trust instead?
What personal information am I sharing?
These are the kinds of questions that help students become more critical users of AI.
The goal is not to scare students away from technology or tell them that they can never use AI.
It is to help them recognise its limitations.
What should schools teach students about AI chatbots?
Schools do not need to wait until a serious incident occurs before having these conversations.
AI literacy should include practical conversations about the way students are already using these tools.
For Years 7–12 students, that could include teaching them:
1. AI can sound like it understands you
Students should understand the difference between an AI system generating a convincing response and a human genuinely understanding their situation.
2. AI advice can be wrong
A chatbot can provide inaccurate or inappropriate information while sounding completely confident.
Students need to develop the habit of questioning the answer rather than simply accepting it because it sounds convincing.
3. Some conversations belong with people
Students should know when an AI chatbot is the wrong place to turn and identify the trusted adults and professional supports they can approach instead.
4. Privacy matters
Students should think carefully before sharing highly personal information with an AI system.
AI literacy and online safety increasingly overlap here. Knowing how a tool works is only part of responsible AI use. Students also need to think about what information they are giving it and what they are using it for.
This conversation needs to start before AI becomes a student's first place to turn
One of the biggest mistakes schools can make is treating AI literacy as a conversation about cheating.
Academic integrity matters, but students are using AI for much more than completing assignments.
They are using it to search for information, brainstorm ideas, solve problems, have conversations and increasingly seek personal or emotional support.
That means AI education in schools needs to include the human side of AI.
Students need to understand not only what AI can do, but also what it cannot do.
That is particularly important when an AI chatbot starts to feel like a friend, confidant or trusted source of advice.
The technology is already part of students' lives. The education needs to catch up.
AI literacy needs to include AI companion safety
For schools, this is one of the reasons AI literacy needs to move beyond prompts and productivity.
Students need practical guidance around AI chatbots, AI companions, critical thinking, privacy and online safety.
They need to be able to recognise when an AI response should be questioned, when a conversation should stop and when it is time to speak to a real person.
That is the kind of practical AI literacy I focus on in my work with schools across Perth and Western Australia.
We need to be having these conversations before AI becomes a student's first place to turn.
This is happening now.
If your school is looking at how to prepare students for the social, emotional and safety challenges that come with AI, explore Paris's High School workshops and programs such as Artificial Connection and Think Before You Prompt. These sessions help students develop practical AI literacy, critical thinking and safer habits around emerging technology.
