A chilling new report from UNICEF, ECPAT, and INTERPOL has pulled back the curtain on a silent digital epidemic: the weaponization of Artificial Intelligence against children.

New data reveals that at least 1.2 million children across 11 countries reported having their images manipulated into sexually explicit “deepfakes” in just the last year. In some regions, the statistics are staggering, representing 1 in 25 children roughly one child in every average-sized classroom.

The crisis is fueled by “nudification” technology; AI tools specifically designed to digitally strip clothing from photos or alter videos to create fabricated, sexually explicit content. While the imagery may be synthetic, UNICEF is sounding the alarm that the trauma is devastatingly real.

“Deepfake abuse is abuse, and there is nothing fake about the harm it causes,” the agency stated in a landmark briefing this week.

The core of the issue lies in a dangerous legal loophole. In many jurisdictions, laws regarding Child Sexual Abuse Material (CSAM) haven’t kept pace with generative AI. Because the images are “synthetic,” some legal frameworks struggle to prosecute the creators or distributors.

UNICEF is now demanding that governments worldwide immediately expand the definition of CSAM to include AI-generated content, criminalizing its creation, possession, and distribution.

The report also shines a light on the tech sector’s inconsistent response. While some AI developers have integrated “safety-by-design” guardrails, many models remain open for exploitation. The risk is highest when these tools are embedded directly into social media platforms, allowing manipulated images to go viral before moderation systems can react.

The UNICEF-backed roadmap for change includes:

• Mandatory Guardrails: AI developers must build systems that make it technically impossible to generate abusive imagery.
• Proactive Detection: Digital platforms must move from “removing after reports” to “preventing circulation” through advanced detection technologies.
• Child-Centric Lawmaking: Governments are urged to stop waiting for the law to “catch up” and act before the next generation of children is targeted.

Perhaps most haunting is the psychological impact on children themselves. The study found that up to two-thirds of children in surveyed countries are actively worried about their likeness being stolen and sexualized by AI.

As the digital landscape evolves, the message from global protection agencies is clear: when a child’s identity is hijacked by an algorithm, the victimization is direct, and the damage to their dignity is permanent.

Source: UNICEF