Update, Tuesday 16 June 2026: Since this blog was published, the government has confirmed plans to ban social media platforms from offering services to under-16s, with protections expected from spring 2027.
Children do not need adults to panic about the internet, but they do need us to keep up.
The debate about children, rapidly evolving technology and online harm has become urgent. However, urgency can narrow our thinking and current debate risks collapsing critical and complex questions of risk, opportunity, agency and rights into the single, oversimplified question: should children under 16 should be banned from social media?
That question matters. A ban might reduce some exposure if it is proportionate, evidence-led and enforceable. But a ban on its own it will not teach a child how to:
- recognise manipulation
- protect themselves from cyberbullying
- crucially assess misinformation
- protect their privacy
- respond to sexual pressure.
The most important question to ask ourselves is: what do children need from us as they grow up in a world that is in part lived online, with maturity and experience?
New research commissioned by Google, based on data collected from more than 6,000 teenagers in the UK shows how teenagers are using AI. The report finds that:
- 67% of teens use AI for creative projects daily or almost daily
- 65% use it for learning more than once a week
- 77% always or often think about the trustworthiness of information when using the internet or AI for learning.
Young people are asking us to understand that online life is already part of how they learn, socialise, seek support and build identity.
And for 13- to 15-year-olds, while AI is largely a learning tool, over 80% say they would turn to parents for problems such as cyberbullying or privacy issues.
As teenagers grow older, these patterns change. By 16 to 18 AI-use is less about homework and more about life management, self-improvement and transitioning into employment.
Older teens are more likely to use AI for language learning, coding and job or internship applications. They are also more sceptical with 52% always verifying trustworthiness and half actively checking for bias.
Even with this growing autonomy, they still need adults, with a quarter worrying their parents lack the skills to recognise AI or fake information.
This transition - and insight from teenagers themselves - is a warning: static control is not good safeguarding.
Naturally, there should be stronger regulation of harmful design features like recommender systems that amplify distressing content, compulsive loops like automatic play functions, weak reporting routes, unsafe default settings and products that treat children as if they are adults.
But even with the best regulation, risks will remain and children will still need adults who can:
- talk to them about what they see
- help them to interpret what feels confusing
- respond calmly when something goes wrong.
This is why we should be thinking seriously about a national behaviour change campaign for trusted adults.
A successful campaign could focus on three things adults can do:
- Ask questions calmly. Trusted adults need prompts that open conversation, such as ‘Show me how this works’ or ‘What do you like about it?’. These questions can help to build trust before a crisis.
- Agree boundaries that grow with the child. Younger teenagers may need safer defaults and limits while older teenagers may need negotiated boundaries and space to practise judgement coupled with support with more complex issues such as AI use or scams.
- Act early and respond without blame. Many children do not ask for help because they fear punishment, embarrassment or losing access to devices. Adults need a simple script that starts with thanking children for sharing and reassuring them they are not in trouble.
Parents have always helped children to navigate friendship, relationships, education and careers. Children growing up online need that same support and increasing trust and autonomy in digital spaces.
The test of success is whether fewer children are harmed and more children can ask for help and children already disadvantaged offline can benefit from technology safely.
Children need adults to keep up, stay close and know what to do when it matters.
Q&A: Jeffrey Demarco answers your questions about the under-16s social media ban
Why shouldn't under-16s have social media?
The concern is not simply that under-16s use social media. The concern is that many services were not designed with children’s safety, development or rights in mind. Children can face so many different risks and harm online from harmful content to bullying; sexual contact risks and image based abuse; and serious issues around manipulative design and algorithmic amplification, even before they have the maturity or support to manage those risks. But the answer cannot only be to keep children out. Children need safer digital environments and not just restricted access to unsafe ones. The real test is whether platforms are required to change the features and business models that create harm.
How does social media affect children under 16?
Children are not a homogenous group and social media affects children in different ways. For some, it can support friendship, creativity, learning, identity and access to information. For others, it can expose them to harmful content, bullying, social comparison, unwanted contact, sexual pressure, misinformation and compulsive design. The impact depends on the child, the platform, the features they use, the support around them and what else is happening in their life. That is why we should avoid basic claims that social media is either all good or all bad. The question is which features create harm, for which children and in which contexts.
Will banning social media for under-16s actually keep them safe?
A ban may reduce some children’s exposure to risks on mainstream platforms but it will not automatically keep children safe. Some children may move to false age or borrowed adult accounts. Equally, they may just spend more time in private messaging or gaming spaces, smaller platforms or they might use VPN supported workarounds. That could make risk less visible to parents, schools and services. A ban can look protective on paper while leaving the deeper digital ecosystem unchanged. So the key question is whether children are actually safer in practice.
If a ban isn't enough on its own, what else needs to happen?
A ban must sit within a much wider child online safety strategy. That means stronger safety-by-design duties, better enforcement of the Online Safety Act, safer recommender systems, limits on stranger contact, high privacy defaults, rapid reporting and removal, action on nudification and sexual deepfakes, safeguards around livestreaming and protections around high-risk AI companion systems. Parents also need practical support and not what may amount to impossible responsibility. This especially matters for children living in poverty, children in unsafe homes and families with fewer resources to understand or manage complex digital rules.
Is it too late once a child already has accounts?
No. It is never too late to make a child safer online. The worst response is panic or punishment, because that can make children hide their online lives. Parents and carers should start with calm, curious conversations and ask questions about:
- what platforms their children use
- who do they speak to
- what feels good to them
- what feels stressful
- where they would go for help when they need it quickly
Then they can review a range of tactics and techniques and develop family social media agreements which includes elements of privacy settings, contact settings, reporting tools, screen habits and risky features, together. The aim is to rebuild trust and make help seeking easier.
Now the ban has been confirmed, what should parents do between now and spring 2027?
Parents should not wait for the law to do all the work. Between now and spring 2027, they can start honest conversations about online life and agree family expectations, check privacy and contact settings, talk about pressure to share images, explain how to block and report and make clear that children will not be punished for asking for help. But parents should also be supported. We know that many families are stretched and digitally overwhelmed and might be dealing with wider pressures. Government and platforms must provide clear guidance alongside the safer defaults and practical tools so that parents are not left to carry this alone.
