Age-appropriate AI practice
How realistic simulations can support judgement without exposing children to real harm.
Research
Questions worth studying with care.
GuardCyber is seeking research partners to examine how children recognise and respond to AI risks, and how schools and families can support them.
Manipulation and over-reliance
How children recognise persuasive AI behaviour and know when to seek human help.
Privacy-conscious learning
How useful group-level guidance might be developed while minimising unnecessary data.
Support over time
How schools and families can revisit learning and understand where further practice may help.
GuardCyber perspectives
Our published thinking, clearly separated from research.
The essays and reflections linked below are written by the GuardCyber team. They are perspectives on child AI safety — not research findings. Articles are currently published in Italian.
These are proposed research areas. GuardCyber does not claim validated measures, proprietary benchmarks or completed research partnerships.
Discuss research collaboration