I have been building online products for almost three decades. Online shops, content sites, marketing platforms, a dating app and an AI companion. I know from the inside how a product is designed to make you use it more, more often, for longer. And I am a father, of a daughter and a son. That combination makes you read every headline about kids and the internet differently.
I am not writing this to scare anyone, or to call for banning everything. I am writing it because, after years of saying “let’s see, let’s talk about it later”, I have concluded that children’s online safety is something we can no longer put off. Not as parents, not as founders, not as a society.
What the risks actually are
The list is longer than I thought ten years ago, and it changes faster than we manage to adapt:
- Grooming and sextortion. Adults who win a child’s trust, then ask for images and use them for blackmail. The child stays silent out of shame, and that silence is exactly what the abuser counts on.
- Cyberbullying. Harassment no longer stops at the school gate. It follows kids into their phones, in the evening, at weekends, in group chats where no adults are present.
- Harmful content and recommendation algorithms. A child who watches one video about dieting may then be served dozens. The algorithm is not malicious, but it optimises for attention, not for the child’s wellbeing.
- Addictive design. Infinite scroll, streaks you are “not allowed” to lose, late-night notifications. I know them well, because they are the same tools the industry uses to boost retention.
- AI chatbots and companions used by minors. Most adults understand they are talking to software. A lonely child may treat a chatbot as a friend or a confidant.
- Deepfakes and image abuse. Today you can fabricate a fake nude from an ordinary profile photo. The victim did nothing wrong, yet suffers as if they had.
- Scams. Fake prizes, “free” game accounts, offers asking for a parent’s card details.
- Data and privacy. Children leave digital traces long before they understand what it means to have a commercial profile built around them.
I have written about some of these before: about how we protect ourselves from deepfakes and about the ethics of AI companions. The difference with children is that they have far fewer ways to defend themselves.
Platforms: safety has to be built in from the start
The biggest responsibility lies with platforms, because they decide what the product looks like. Safety is not a button you add after a scandal; it is a design decision. Here is what I think should be standard:
- real age assurance, done in a way that respects privacy, without demanding an ID card from everyone who opens an app;
- minors’ accounts private by default, no messages from strangers, not easily discoverable;
- no profiling-based advertising aimed at minors;
- simple reporting and a fast response, especially for sexual abuse and intimate images shared without consent;
- no streak mechanics or night-time notifications on children’s accounts.
Europe is already moving this way. The EU Digital Services Act (DSA) requires online platforms to ensure a high level of privacy, safety and security for minors, and bans profiling-based advertising targeted at them. In July 2025 the European Commission published guidelines on the protection of minors under the DSA and presented a prototype age-verification app, piloted in several member states. In the UK, the child-safety duties of the Online Safety Act, including age checks for some types of content, also started applying in July 2025. Australia went further still: since 10 December 2025, under-16s can no longer hold social media accounts.
What I do as a founder who builds products
It is easy to point at the giants. It is harder to look at your own products. My rule is simple: if a product is not meant for children, it has to behave that way. Slide, the dating app, is adults-only and built on verified profiles. Desirex, the AI companion, is adults-only too. I have written in more detail about safety in dating apps: verification, moderation, easy reporting, and real people who actually read the reports.
None of this is perfect. But there is a huge difference between a product that seriously tries to keep minors out and one that just ticks an “I am over 18” box. I think every founder should ask the uncomfortable question: what would happen if a 13-year-old landed in my app tomorrow?
Parents: more conversation, not just control
Parental control apps have their place, especially for younger kids. But no filter replaces a relationship in which your child tells you when something feels wrong. A child afraid of losing their phone will not mention the strange message they got. One who knows they will not be punished for someone else’s wrongdoing is far more likely to speak up.
Rules agreed together, rather than imposed, help. Keeping phones and tablets out of the bedroom at night helps. Knowing which apps your kids use, and trying them yourself, helps. And, perhaps hardest of all, being an example helps. You cannot ask a teenager to put the phone away at dinner while you answer emails between bites. I have written about burnout in the online generation, and I do not think it is only a children’s problem.
Schools and the state
Digital education is not just knowing how to make a slideshow. It means recognising manipulation, understanding what an algorithm is, and knowing what to do if someone threatens you with an image. This can be taught, and it should be taught from the early school years, not in one ceremonial lesson a year.
As for the state, in Romania and at EU level, a good share of the rules already exists. What is often missing is enforcement. A law nobody checks is just a good intention. Regulators need people, technical skills and the courage to fine the big platforms too, not only the small ones.
The balance: protection without mass surveillance
I do not want a world in which every adult scans their face to read the news, or one in which everyone’s messages are read pre-emptively. Protecting children must not become a pretext for blanket surveillance. Age checks can be done with tools that confirm only “over 18”, without learning who you are. Total bans, for their part, can push kids towards even less safe corners of the internet.
The right path is in the middle: products that are safe by default, rules that are enforced, involved parents, and children who gradually learn to find their own way.
A short checklist for parents
- Talk often, briefly and without drama about what your child does online. Ask what they enjoyed, not only what went wrong.
- Tell them clearly: if anyone asks for pictures or threatens them, they come to you and they will not be punished.
- Set accounts to private together and check who can message them.
- Keep phones and tablets out of the bedroom at night.
- Install and try the apps they use yourself.
- Switch off unnecessary notifications and talk about streaks and infinite scroll.
- Explain what a deepfake is, and that a fake image is never the victim’s fault.
- Report abusive content on the platform and, if needed, to the police.
- Be the example you want them to see.
Children’s online safety is not one actor’s job. It is shared between those who build the products, those who make the rules and those who sit at the dinner table with their kids every evening. How do you handle this at home? Which rule has helped you most?