When Their Friends Are All Online

Start by not making the first mistake

The first mistake is telling a teenager that their online friends are not real friends. It is almost never true and it always costs the same thing: they stop telling you about them. After that you are watching a part of their life through a keyhole, which is worse for their safety, not better.

For some teenagers this is not a substitute for anything. Think of a teenager who lives twenty miles from the nearest person their age. Or one whose interest has four other people in the county. Or one who finds a loud room genuinely hard. For all three, online is not the lesser version. It is where the friendship is.

What the evidence actually says, including the part that cuts the other way

This is an unsettled question and the honest answer is that the field disagrees. Here is both sides, as fairly as we can put them.

The case that the panic overshoots. The National Academies reviewed the published literature in December 2023. Its review "did not support the conclusion that social media causes changes in adolescent health at the population level". It added that the available research linking social media to health "shows small effects and weak associations". Writing in Nature in March 2024, Candice Odgers made a similar point. The claim that these technologies are rewiring children's brains and causing an epidemic of mental illness "is not supported by science", she wrote.

On friendship in particular, teenagers' own reports skew positive. Pew surveyed 1,391 US teenagers aged 13 to 17 in autumn 2024, and published the results in April 2025. In it, 74 percent said what they see on social media makes them feel more connected to what is going on in their friends' lives. Thirty percent said it helps their friendships. Seven percent said it hurts them.

And loneliness does not track access the way the simple story predicts. A 2025 study in the Journal of Adolescent Health covered 38 countries and over a million adolescents, from 2000 to 2022. It tested six societal factors against rising loneliness. None was directly associated with it. More internet access went with less loneliness, not more.

The case on the other side, which is not weak. Studies that switch platforms off for a period tend to find benefits. The pattern of girls faring worse than boys turns up across many separate studies. And mental health measures fell in many countries at around the same time, which is hard to explain with causes specific to one country.

We are not going to tell you who is right, because nobody knows yet. What we will say is that the better-evidenced harms are displacement harms, sleep most of all, rather than direct damage to friendship itself.

The statistics you are being shown are old

If you searched this question before landing here, the numbers you saw about teenagers meeting friends online most likely came from a survey fielded in 2015. They are eleven years old and are still circulating undated, including on a page run by a company that sells monitoring software.

That is worth knowing about a page telling you how worried to be. A number with no date on it is not a fact about your teenager's year.

The two warning signs that matter most

These are not judgement calls. Treat them as rules.

Someone who is cagey about their age, or who turns out to be an adult. Age is the easiest thing in the world to answer. Dodging it, changing it, or getting vague when asked is a signal on its own. If an online friend is evasive about their age, or turns out to be an adult, the response is to block them and tell an adult you trust. That is the whole rule and there is no version of it that requires being certain first.

Anyone who asks for photos. Any request for pictures, of any kind, however it is framed, and however long you have been talking. Block them and tell an adult you trust. It does not matter whether the request seemed harmless, whether they sent one first, or whether anything has already been sent. Telling someone is the step that stops it, and a teenager who has already sent something needs to know that the trouble they are afraid of is far smaller than the trouble of handling it alone.

Both rules work better if they are stated once, plainly, before anything happens, and stated as rules about other people's behaviour rather than tests of your teenager's judgement.

The other two narrow concerns

Sleep. This is the one with the clearest evidence behind it and it is easy to act on. The phone charges somewhere other than the bedroom. That is a rule about sleep, not a punishment about friends, and it is worth saying so out loud.

One friendship that has swallowed the others. Not because online friendship is dangerous, but because any single relationship that is on constantly, demands immediate replies and reacts badly to time apart is a heavy thing at fifteen. That pattern is worth noticing whether it is happening on a screen or at school.

Neither of these requires ending anything. Both are conditions you can change.

What this comes down to

Your leverage here is the same as everywhere else in this hub. You do not get to choose who your teenager cares about. You do get to set the conditions: where the phone sleeps, what the two rules are, and whether you are someone who can be told things.

The last of those is the one that actually protects them, and it is the one most easily spent on being dismissive about people you have never met.