A summary of UNICEF’s February 2026 Issue Brief on Artificial Intelligence and Child Sexual Abuse and Exploitation

In February 2026, UNICEF released an issue brief addressing the intersection of artificial intelligence and child sexual abuse and exploitation. The report clearly defines the scope of this growing problem and emphasizes the urgent need for coordinated action from governments, technology companies, and other stakeholders.

The brief highlights that open-source software capable of generating hyper-realistic images and videos is now easily accessible and can be deployed on virtually any device. Because these tools are widely available, they can be misused by those with harmful intentions to create content that endangers children.

One of the most alarming points from the issue was the manipulation and creation of images into sexually explicit deepfakes of children. In the report it was mentioned that children are aware of this threat and these images can cause social isolation, stigma and long-term emotional drama. One in 25 children are victims in images or videos where their faces have been misused.

UNICEF identified several urgent priorities

Even thought Finland in the report it is not explicitly mentioned and there is no direct data collected from Finland it is important to be ready and pay attention to the priorities:

For families and schools: Parents and schools need to educate children to be aware of this type of harm. Educators and professional working with children need to have training to respond to the deepfakes and tools created for editing pictures and videos. They also need training about the direct and indirect psychological impact on the children.

For governments: National laws must be updated to explicitly cover AI-generated CSAM. Legal frameworks need to align with international standards and ensure real remedies for child victims.

For the tech industry: Companies must conduct child rights impact assessments and be transparent about how their products affect children. Every actor in the AI supply chain — from dataset providers to model developers — should embed safety into their systems from the start, including pre-release safety testing for open-source models.

Lesson to be learned

The brief makes clear that teaching children to be careful about what they share online is no longer enough on its own. A child’s body, identity, and reputation can now be violated without their knowledge or participation. Addressing this reality requires coordinated action: stronger laws, more responsible technology development, better-equipped professionals, and informed communities.