Keeping Children Safe in the Age of AI
Mrs Houda Benhamouda, Head of Computing and Digital Literacy, Loughborough Schools Foundation
Artificial intelligence is woven into the everyday digital experiences of young people. From homework support tools to social media algorithms and increasingly sophisticated content creation, AI is shaping how children learn, communicate, and understand the world. The challenge for schools and parents is not to resist this change, but to guide children through it safely, confidently, and critically.
At Loughborough Schools Foundation (LSF), we believe that safeguarding in the age of AI is not about fear, it is about equipping children with the awareness, habits, and resilience they need to navigate a rapidly evolving digital landscape.
A new kind of digital literacy
Traditional online safety has focused on limiting exposure: blocking harmful websites or restricting screen time. While these remain important, AI introduces a more complex challenge, one where not everything harmful is obviously inappropriate.
AI-generated content can appear polished, convincing, and authoritative, even when it is entirely incorrect. Children must now learn to question not just where information comes from, but whether it is true at all. AI systems can produce plausible yet fabricated facts, which makes critical thinking an essential life skill for today’s children, rather than an academic exercise.
Equally, the rise of deepfake technology means children can no longer assume that seeing or hearing something online makes it real. Voice notes, images, or videos may convincingly mimic trusted individuals, creating new risks around manipulation and deception.
As a response to this, parents should encourage digital literacy skills that also include digital scepticism.
From protection to empowerment
Safeguarding is often framed as protection, and rightly so. Schools implement filtering systems, device management, and strict acceptable use policies to reduce exposure to harm.
However, no technical system can fully eliminate risk. Children will inevitably encounter unfiltered environments beyond school on personal devices, social platforms, or through peers. This is why it is important that we ensure children understand that not everything online is real and that AI outputs are not always reliable.
Simple frameworks such as pausing before sharing, investigating sources, and cross-checking claims help children develop habits that will protect them far beyond the classroom.
AI as a tool, not a shortcut
AI has enormous educational potential. Used well, it can break down complex ideas, support revision, and encourage independent learning. But used poorly, it can undermine the very skills education seeks to develop.
One of the most important messages we share with students is that AI should be a thinking partner, not a replacement for thinking.
When children rely on AI to complete work for them, they lose the opportunity to develop understanding, resilience, and problem-solving ability. More importantly, they may not recognise when the information they are given is flawed or biased.
Alongside teachers, parents must encourage pupils to use AI to generate ideas, not final answers; check outputs against trusted sources and to question inconsistencies or surprising claims. This helps ensure that AI enhances learning rather than diminishing it.
The hidden risks: social and emotional impact
While much attention is given to misinformation and technical risks, the emotional dimension of online life remains just as critical.
AI-driven platforms can amplify content designed to engage, often by triggering strong emotional responses. Combined with unmoderated spaces such as livestreams or comment sections, children may encounter harmful language, unrealistic expectations, or toxic interactions.
Warning signs of online harm are not always obvious, but may include:
- Increased secrecy around devices
- Sudden changes in online behaviour
- Withdrawal, anxiety, or distress after screen use
- Reluctance to discuss online interactions
Recognising these signs early allows for supportive intervention and highlights the importance of strong relationships over strict surveillance.
The role of parents: conversation over control
Parents play a crucial role in reinforcing these messages and providing the context in which children apply them.
One of the most consistent findings in online safety research, and in our own experience, is that the most effective safeguard is not technology, but communication.
Children are far more likely to seek help if they feel they will be listened to, rather than judged or punished. As part of our knowledge-sharing with parents as LSF, we encourage parents to engage in the following:
- Talk regularly about what children are doing online
- Show curiosity rather than suspicion
- Agree clear boundaries together
- Reassure children they can report concerns without fear
My key message to parents is reassuringly simple: you do not need to be a technology expert to keep your child safe. What matters most is staying engaged, asking questions, and keeping communication open.
To navigate this new landscape, I recommend that parents focus on three core principles:
- Protect – Set clear boundaries around device use, enable filtering where appropriate, and understand the platforms your child uses.
- Detect – Stay alert to behavioural changes and take concerns seriously without overreacting.
- Empower – Encourage children to think critically, question information, and use AI responsibly.
A shared responsibility
Keeping children safe in the age of AI is not the responsibility of schools or parents alone. I believe that it requires a partnership built on shared understanding and consistent messaging.
Schools must continue to maintain robust safeguarding systems, teach digital literacy and parents should model responsible AI use themselves.
AI will continue to evolve, both rapidly and unpredictably. But the fundamentals of keeping children safe remain unchanged. Strong relationships, open conversations, and critical thinking will always be more powerful than any piece of software.
