[Survey Results] Over Half Aware of 'Sexual Deepfakes,' Victim Recognition Also Increasing. Countermeasures Against AI-Generated Child Human Rights Violations Include Legal Regulations, Industry Regulations, and Improving Children's AI Literacy.

Child Fund Japan (Suginami-ku, Tokyo; Secretary-General: Katsuhiko Takeda), a specified non-profit organization, conducted a national awareness survey on "Generative AI and Child Human Rights Violations" targeting men and women aged 15-79 across Japan in 2026.

The survey results revealed an increase in people aware of sexual deepfakes, and a growing demand for legal and industry regulations to address online child rights violations, as well as for children to acquire self-defense skills such as AI literacy.

---

Protecting Children's Rights from Generative AI and Deepfakes! Advocacy Activities from the Perspective of Children and Youth.

With the rapid evolution and widespread adoption of generative AI, large volumes of images can now be instantly created and disseminated. In recent times, issues such as Child Sexual Abuse Material (CSAM) and sexual deepfakes generated by AI are spreading, necessitating legal frameworks and increased public awareness.

Child Fund Japan has previously undertaken initiatives to protect children from sexual exploitation, including surveys on grooming targeting youth and creating animated videos to warn children and young people about grooming. Furthermore, in April 2024, they held a symposium titled "The Threat of Generative AI to Children's Rights," and in December, an online seminar to learn about pioneering initiatives overseas. They also launched a working group of experts to propose new legislation to regulate AI-generated CSAM.

As part of these efforts, Child Fund Japan conducted a national awareness survey on the potential for generative AI to violate children's human rights and on possible countermeasures.

Key findings from the survey are as follows:

■ Detailed Survey Results (Partial)

*In the survey, each question was asked after providing definitions of terms (see footnotes).

Q: Are you aware of "sexual deepfakes"?

52.7% responded "Heard of it," approximately half.

Q: Please indicate your thoughts on how to address the issue of AI-generated CSAM involving children.

Regarding responses to the issue of AI-generated CSAM involving children, "The government should create laws to regulate it" (46.8%) and "SNS and app providers should impose certain restrictions on children's use of services" (43.7%) were the most common. These were followed by "The AI-related industry should impose technical regulations" (35.8%), "Companies should restrict children's use of devices (smartphones, etc.)" (34.9%), "Children themselves should enhance their AI literacy to develop self-defense skills" (34.0%), and "The government should regulate it through an AI strategy" (33.1%).

Based on the results of this survey, the research team makes the following recommendations:

Child Fund Japan will continue its activities, including policy recommendations and further research on children and youth, to protect children's rights.

The answers to each question, analysis, recommendations, and survey questionnaire are available at:

https://www.childfund.or.jp/blog/20260330survey

■ Request for Credit Notation when Citing/Reproducing

When citing or reproducing the survey results, please include the credit: "Survey by Child Fund Japan, a Specified Non-Profit Organization."

(Footnotes)

With the development of generative AI (artificial intelligence), content that sexually objectifies children*1 (CSAM: Child Sexual Abuse Material)*2 can be easily created and uploaded to SNS for dissemination.

Furthermore, generative AI can easily create fake sexual content of children from images of "real children." It can also easily create sexual content of "fictional children" who do not exist.

Additionally, determining whether the depicted child is real has become difficult due to the advancement of generative AI.

*1 In this context, "children" refers to "individuals under 18 years of age (minors)."

*2 "Content" refers to expressions in "images, videos, audio, etc." "CSAM: Child Sexual Abuse Material" is an abbreviation for Child Sexual Abuse Material, some of which are not currently subject to legal regulation.

"Sexual deepfake" refers to content created using generative AI to create a person, often sexual content created by combining the face of a real person with a fake naked body. These also fall under "CSAM."

Apps and services that create such CSAM also exist, and children are also falling victim. They can also be used for bullying and harassment, and in some cases, the perpetrators are children.

~About Child Fund Japan~

An international cooperation NGO that has been working since 1975, primarily in Asia, to support the healthy growth of children living in poverty and the self-reliance of their families and communities. Centered in the Philippines, Nepal, and Sri Lanka, they continue to support children through sponsorship programs (programs that support children's growth through letter exchanges with local children). They also collaborate with 10 member organizations of Child Fund to deliver aid to 36 million people in 66 countries worldwide.
They engage in activities that contribute to achieving SDGs Goals 1, 3, 4, 5, and 16, with a particular focus on "child protection" through awareness and advocacy activities to achieve Goal 16.2, "End abuse, exploitation, trafficking and all forms of violence against and torture of children."