26/04/2026
We increasingly rely on AI. But how far should that reliance go?
Let’s take a look at this image.
In one frame, an AI tool confidently labels a toxic mushroom as safe to eat.
In the next, it offers a polite apology. But in the real world, an apology does not undo harm.
It is a stark reminder of how decisions are shaped in environments where speed and confidence can hide dangerous uncertainty.
The illusion of Authority
Children and young people are growing up in a world where AI is always present, responsive, and often seen as authoritative.
When something sounds fluent and assured, it is easy to assume it is correct.
But children are still developing the ability to question, verify, and spot subtle inaccuracies. The real risk is not just misinformation. It is misplaced trust.
And that trust can have serious consequences:
• Health and Safety: medical or nutritional advice taken at face value
• Physical Risk: exposure to harmful challenges or unsafe recommendations
• Emotional Development: distorted views of identity, relationships, and self-worth
Moving Beyond user Responsibility
This is not simply about how users behave. It is about how systems are designed.
If AI is built to sound certain, it must also be built to signal uncertainty, especially when children are involved.
What needs to change?
✅ Safety by Design: wellbeing must be embedded from the outset, not added after harm occurs
✅ Transparent Uncertainty: clear signals of limits should be standard, not optional
✅ Critical Literacy: children need support to question and think, not just access tools
✅ Collective Guardrails: platforms, educators, and parents must work together to create safer environments
AI can be a powerful co-pilot. But children should never be left to navigate it alone.
The true measure of technology is not how intelligent it appears, but how well it protects the rights, safety, and wellbeing of every child.
Illustration: Unidentified Creator (please write the name of the creator, if you know)