A UNICEF study reveals that 60% of online sexual abuse of children occurs on social media platforms.

      A UNICEF study published on Thursday revealed that around 20 million children experienced online sexual exploitation or abuse within a single year across 21 countries. This equates to approximately one in five children aged 12 to 17 who use the internet in the surveyed nations.

      The study titled “Through Children’s Eyes: How Digital Technologies Enable Child Sexual Abuse,” was created as part of the Disrupting Harm project, which UNICEF runs alongside ECPAT International and INTERPOL, and has been reported by Reuters. It is considered one of the largest studies of its kind.

      The strength of the findings lies in the methodology used. Researchers conducted nationally representative surveys involving about 21,000 internet-using children across Africa, Asia, Latin America, and Eastern Europe between 2020 and 2025, along with 100 in-depth interviews with survivors.

      Breaking down the estimates, over 15 million children were exposed to unwanted sexual content, more than nine million faced pressure for sexual conversations or were urged to share images, and four million had intimate images distributed without their consent. These categories can overlap, meaning a single child may have experienced multiple instances.

      Nearly 60% of these incidents occurred on social media platforms, with Facebook, Instagram, Snapchat, TikTok, and WhatsApp highlighted in the study. An additional 14% occurred within online games.

      This concentration of incidents is significant for policy discussions, as the platforms responsible for the majority of these harms are the same ones confronting age verification requirements in Europe and bans for under-16s in various countries. The study provides regulators with a statistic to reference instead of relying on anecdotes.

      Generative AI has introduced a new dimension to the issue. In nine of the countries studied, an estimated 1.1 million children reported that sexual images of them had been created using AI—a concern that did not substantially exist when the initial fieldwork started in 2020.

      This issue accompanies warnings to parents regarding AI-generated abusive imagery. It also complicates detection efforts, as systems designed for identifying known materials struggle with entirely new images.

      The nine countries contributing to the AI figure align with the second round of fieldwork conducted in 2024 and 2025. The twelve countries surveyed four years prior were not included in this latest round, leaving an absence of baseline data to evaluate the rate of growth in this problem.

      Data on perpetrators challenges the common “stranger danger” narrative often emphasized in parental guidance. In 57% of cases, the person was known to the child—either a peer, partner, friend, or family member—while 38% involved individuals whom the child only knew online with fake accounts.

      The repercussions of this abuse are severe and often go unacknowledged. Children subjected to such experiences reported suicidal thoughts and self-harm rates approximately four times higher than their peers, along with increased anxiety.

      Almost none of these cases are reported to authorities. Fewer than 1% were brought to the attention of law enforcement or support services, indicating that official statistics reflecting enforcement priorities capture only a fraction of the actual problem.

      UNICEF recommends a focus on product design rather than solely on parental vigilance. John Livingstone, head of digital policy at UNICEF Australia, stated, “Platforms must build safety into products from the outset, rather than relying on children and parents to manage risks after harm occurs.”

      The survey was conducted in two phases. Countries including Cambodia, Ethiopia, Indonesia, Kenya, Malaysia, Mozambique, Namibia, the Philippines, Tanzania, Thailand, Uganda, and Vietnam were studied in 2020 and 2021, while Armenia, Brazil, Colombia, the Dominican Republic, Mexico, Montenegro, North Macedonia, Pakistan, and Serbia were added in 2024 and 2025.

      However, the exclusion of the United States, the United Kingdom, and EU member states from the sample limits the scope of the estimate. It represents a floor of 21 countries rather than a global total, and UNICEF has not extrapolated the findings beyond this group.

      These findings emerge as regulators examine whether platforms are adequately addressing issues related to child sexual abuse and sexual extortion. Australia’s eSafety commissioner has already identified significant deficiencies in how large companies respond to such incidents. UNICEF has also found that children are adopting AI tools at approximately three times the rate of adults.