What must change Taken together, these findings tell a clear story. The harm is widespread and it reaches children through everyday platforms; children’s trust is exploited and harm is enabled by platform design choices; and when children seek help, the systems built to protect them are not seen as viable options. Children often respond with remarkable resourcefulness – blocking perpetrators, avoiding certain online spaces and protecting each other – but their resilience is not a substitute for effective protection. That they must cope so much alone is itself evidence of societies and systems failing them. Children should not have to fight so hard to be safe and supported. The current situation is failing children very badly. The answer is not in simply restricting access to certain sites, displacing children and abusers to other parts of the internet or creating more reasons to hide problems. The evidence points to clear, achievable action by those with the power to act. The answer lies in addressing how platforms are designed by companies, how online spaces are regulated by governments and the creation of societies and systems that make it much easier for children to recognize and speak up about abuse, reach out for help and get protection and redress, so no child has to face this alone. Technology companies must design for safety, not maximum engagement. Platforms should establish privacy-by-default protocols for children, restrict unsolicited contact from adults, ensure recommender systems neither surface children to potential perpetrators nor promote sexualizing content and proactively detect abuse material, including AIgenerated content. Governments and regulators must make safety enforceable. Governments should regulate for safety-by-design, equip regulators with the technical expertise, resources and powers to hold industry to account, and update laws to cover grooming, sexual extortion and AIgenerated abuse material. Through Children’s Eyes Executive summary 4