NEW YORK — Nearly one in five internet-using children aged 12 to 17 in 21 countries experienced at least one form of technology-facilitated sexual exploitation or abuse during a single year, according to a UNICEF report released on 3 September.
The estimate represents about 20 million children in the countries studied. More than 15 million encountered unwanted sexual content, nine million were pressured into sexual conversations or image-sharing, and four million had sexual images distributed without consent.
Almost 60% of incidents occurred on mainstream social platforms, including Facebook, Instagram, Snapchat, TikTok and WhatsApp. Online games accounted for 14%. The evidence challenges the idea that danger lives only on obscure websites.
It also challenges the stranger-danger model. In 57% of cases the child knew the perpetrator as a peer, romantic partner, friend or relative. Thirty-eight per cent first encountered the person online.
UNICEF combined nationally representative surveys of about 21,000 internet-using children with in-depth interviews involving 100 young people who had experienced abuse. Data came from two research rounds, primarily in 2020–21 and 2024–25, across countries in Africa, Asia, Europe and Latin America.
The results represent internet-using children in those countries, not every child worldwide. Because disclosure is difficult, UNICEF says the true scale is probably higher. More than 40% of experiences were shared with nobody, and fewer than 1% were reported to police, social workers or helplines.
Silence should not be interpreted as consent or absence of harm. Children exposed to the abuse were four times more likely to report self-harm or suicidal thoughts and behaviour, although the survey cannot prove the abuse caused every mental-health outcome.
Gender belongs inside the response. The report calls for action against norms that normalise sexualisation of girls and stigmatise survivors. Girls can face additional reputational punishment when an image circulates, while boys and gender-diverse children may remain silent because abuse conflicts with social expectations about masculinity or identity.
Platforms should restrict unsolicited adult contact, make minors’ accounts private by default, interrupt rapid forwarding of intimate material and provide reporting that preserves evidence without forcing a child to repeatedly view it. AI-generated sexual images must be treated as abuse even when no camera was involved.
Parents and schools need practical guidance, but surveillance of children is not a safety design. The burden belongs with companies that construct messaging, recommendations and virality; governments that enforce law; and services able to respond without blame.
Twenty million is not a warning about children spending too much time online. It is evidence that products central to childhood have been allowed to externalise the cost of safety onto their youngest users.
Join the discussion