For some children, going online is where they first find the confidence to speak.
I know this because I was one of them. As someone with autism and ADHD, I often felt like an alien outside the online world. Through gaming and online communities, I learned to express myself, connect with people and discover what I could contribute. Without those experiences, I would not have become the person I am today.
That is what I carry into the debate about restricting children’s access to social media: concern for their safety, alongside an understanding of what they might lose.
On 16 September, European Commission President Ursula von der Leyen outlined an EU Kids Act: no social media below 13, parent-supervised “mini accounts” for 13- and 14-year-olds, and mandatory safer design for ages 15–17. The detailed proposal was scheduled for the following day. This was an announcement, not an EU-wide ban already in force. European Commission address
News coverage also reported a one-hour daily limit for the mini accounts. That number deserves scrutiny. What evidence establishes one hour as the right boundary, for which activities and which children? WIRED, 16 September
I worry that MEPs and other policymakers underestimate how early digital culture becomes part of someone’s identity. Friendships, humour, creativity and belonging develop there. A restriction can remove a source of harm; it can also remove a community.
These proposals concern social-media access, rather than the whole internet. Gaming and social media are distinct, and evidence about one cannot automatically justify the other. But their communities overlap. Policy needs to understand those connections.
The positive evidence deserves space in this discussion. A 2024 Pew Research Center report found that 47% of American teenage gamers had made an online friend through gaming, while 72% played partly to spend time with others. These are teenagers’ reported experiences, not proof that gaming improves everyone’s social life. They nevertheless challenge the assumption that time spent gaming is necessarily time spent alone. Pew Research Center
There is also research behind the familiar claim that gaming improves “reflexes”, although that word is imprecise. Action-game experiments have found faster perceptual decisions without sacrificing accuracy. A broader meta-analysis found benefits in attention and spatial cognition, while warning that publication bias inflated estimates. These findings concern particular games and measured skills; they do not establish that every game improves intelligence or that more hours produce greater benefits. Green and colleagues, Bediou and colleagues
The risks deserve equal honesty. In Pew’s survey, 41% of teenage gamers said gaming had harmed their sleep, and 41% reported being called offensive names while playing. Friendship and harm can exist in the same environment. That makes the quality of the environment, its moderation and its place in a child’s day essential questions. Pew Research Center
My experience cannot represent every autistic person or everyone with ADHD. It does show why we should ask what function an online space serves before removing it. If it is where someone feels understood, their wellbeing cannot be measured only by the minutes they spend there.
I would also like researchers to test the device question more directly. Does using the same platform on a desktop computer have the same effects as carrying it everywhere on a smartphone? How much comes from the content, the recommendations, notifications, bedtime use or constant availability?
A 2025 randomised trial offers a useful clue. Researchers blocked smartphone internet access for two weeks while leaving calls, texts and internet access through computers available. They found improvements in mental health, wellbeing and sustained attention. However, the participants were adults, compliance was limited, and the experiment did not compare identical social-media use across devices. It supports investigating constant mobile access; it does not prove that desktop social media is harmless for children. PNAS Nexus study, research team’s university report
A good comparison would track the same activities across devices, account for existing wellbeing and family circumstances, and measure sleep, friendships, learning and exposure to harm. “Screen time” alone cannot tell us enough.
Australia demonstrates why implementation matters. Its under-16 account restrictions took effect on 10 December 2025. In March, the eSafety Commissioner reported millions of accounts removed or restricted, alongside substantial continued access and weaknesses in age checks. Account removals are evidence of enforcement activity; they do not, by themselves, establish improved wellbeing. eSafety compliance report
Workarounds included fake accounts and VPNs; around 3% reported VPN use in one early study. The Guardian, 24 June
VPNs were a minority workaround. Early circumvention also does not settle whether restrictions can deliver longer-term benefits. We need to measure outcomes and unintended consequences.
From a cybersecurity perspective, I want those consequences examined before restrictions expand. Will children move to poorly moderated services? Will families encounter fake verification pages? What personal information will age checks collect, who can access it, and when will it be deleted? These are design questions that belong inside a child-protection policy.
I am open to limits, and would assess overall recreational screen time separately from access to particular platforms. For school-age children below eight, my starting point would be a maximum of one hour daily. Between eight and twelve, I would consider a ceiling of three hours, with breaks and parental involvement. These are personal proposals to investigate, not scientifically established safe allowances or daily targets. They should be revised if evidence shows worse outcomes.
Infants and preschoolers require separate consideration. WHO guidance recommends no sedentary screen time for children below two and no more than one hour for ages two to four, with less preferred. WHO guidance
For older children, any arrangement must protect sleep, movement, schoolwork and relationships. Required schoolwork should be distinguished from recreational use, and teenagers should gain increasing responsibility. A creative project, cooperative game and algorithmic feed deserve different consideration, even when they occupy the same screen.
Alongside boundaries, schools should teach AI literacy and cybersecurity through age-appropriate lessons: checking sources, recognising phishing, protecting accounts, understanding consent and privacy, identifying manipulated images, and knowing when AI produces confident nonsense. Children should learn how recommendation systems shape attention and how platforms earn money from it. Teachers and parents need support too.
Education must accompany enforceable responsibilities for platforms: effective reporting, safer defaults, limits on manipulative design and protection from unwanted contact. Children cannot be expected to outsmart systems deliberately built to hold their attention.
I want young people, including neurodivergent users, involved in designing and evaluating these policies. Ask them what hurts, what helps and what they would lose. Measure safety alongside belonging and capability.
Digitalisation is already part of the world they are growing into. Our responsibility is to help them participate with knowledge, support and meaningful protection. For a child who feels like an alien everywhere else, finding their people online can be part of growing up.