Edition 2 · 01 / 08
What Do We Mean by Social Media?
A blanket label conceals important differences between public feeds, video libraries, messaging services and school communities.
I am fully sympathetic to efforts to protect children online. The harms are real, and leaving families to manage them alone is not an adequate response. But a blanket ban on social media is not, in my view, a sufficiently precise corrective action.
The first difficulty is definitional. Does social media mean Facebook and public feeds? Does it include YouTube, group messaging, a moderated community for disabled young people or a forum created by a school for its pupils? These services have different functions, risks and benefits. A rule built around an unstable label may prohibit valuable uses while leaving the underlying design problem intact.
The category is not the risk
A public feed can expose a child to strangers and harmful material. A private class forum can help pupils collaborate. A video library can provide entertainment, instruction or repeated examples of speech. Even within one service, the experience may differ according to account settings, age, history and whether a parent participates.
This does not make broad services harmless. It means that policy should identify the relevant functions: public distribution, private communication, profiling, autoplay, recommender systems, contact from unknown users and commercial targeting. Once the function is named, a proportionate control becomes easier to design and evaluate.
The UK’s Online Safety Act guidance from Ofcom already takes a risk based approach to services likely to be accessed by children. The ICO Children’s Code places the child’s best interests at the centre of design and addresses profiling and nudges. The useful question is how these duties apply to the sequence a recommender constructs, as well as the eligibility of each item in it.
Children are not interchangeable
Age matters, but it is not the only difference. Children vary in maturity, disability, communication, interests, family context and the reasons they use a service. Repetition may be dull for one child and useful or regulating for another. Variety may open a new interest or cause distress.
A policy concerned with children’s interests must therefore preserve agency and accommodate difference. It should not assume that one approved experience will serve everyone.
Start with the decision being made
Instead of asking whether a product belongs inside the social media category, ask what decisions it makes about a child:
- What information does it infer?
- What material is eligible to appear?
- What determines the order?
- Does the system create natural stopping points?
- Can the child or parent broaden, reset or escape an inferred interest?
- What evidence shows that the experience serves the child?
That approach reaches the actual mechanism. It can regulate a risky function wherever it appears without pretending that every digital community is the same.