Nearly 1 in 4 Vulnerable Children Use AI Chatbots Because They Have No One Else to Talk To
Internet Matters surveyed 1,000 children and 2,000 parents across the UK and found that 71% of vulnerable children use AI chatbots, compared with 62% of other children. Among vulnerable children, nearly one in four say they use a chatbot because they have no one else to talk to.
The number to hold: 23%. That is the share of vulnerable children (those with special educational needs or a condition needing professional support) who say a chatbot is there because nobody else is. Across all children who use chatbots, the figure is 12%, or about one in eight.
Source: Internet Matters, "Me, Myself and AI" (UK survey of 1,000 children aged 9 to 17 and 2,000 parents). The report was published in 2025 and is UK-based; we are citing it from the organisation's published summary and press coverage, not an independent analysis. It remains one of the clearest looks at which children lean hardest on chatbots.
The Children Who Need People Most Are Reaching for Software
Most AI-safety talk assumes a typical teen with a typical friend group. This research points somewhere more uncomfortable: the children with the fewest places to turn are the most likely to turn to a chatbot, and the most likely to feel it is a real friend.
A chatbot never gets tired, never judges, never tells a teacher and never says "not now." For a child who finds conversation hard, or who feels left out, that is a powerful pull. It is also exactly what makes it risky: it is built to keep replying, not to notice when a child needs a real person.
"I don't have anyone else to talk to." — the sentence the researchers heard from children, and the one parents most want to hear first.
Advice From a Friend That Is Not a Person
Two in five children who use chatbots have no concerns about following the advice they get, and a further 36% are not sure whether they should be concerned. The researchers found the trust was higher still among vulnerable children. Chatbots can be confidently wrong, and a child with little else to compare it to may not know.
Heavy chatbot use does not look like trouble from the outside. A child sitting quietly with a phone looks calm. The warning sign is not the screen, it is a child whose main confidant has stopped being a person.
What to Do This Week
The goal is not to take the chatbot away. It is to make sure it is never the only one listening.
Start a free 14-day Blaick trial — no credit card required.
Blaick helps you see how your child uses AI and flags worrying patterns, so you can step in with a conversation, not a crackdown.
Start free trial →