Artificial intelligence is changing the way people learn, communicate and create content. Children and young people are already using AI tools at school, at home and on social media. These technologies can offer many benefits. However, they can also be used in harmful ways. One of the most serious emerging risks is the use of artificial intelligence to create false sexual images of children. A new guide for parents and carers, produced by CEOP Education, the National Crime Agency and the Internet Watch Foundation, explains how this form of abuse can happen and what adults can do to protect children. The guide focuses on a difficult but increasingly important reality: an ordinary photograph of a child can be taken from the internet and changed into a sexualised image without the child’s knowledge or consent. 

How Can AI Be Used to Harm Children?

Artificial intelligence can create realistic pictures and videos or change existing ones. This means that someone can take a normal photograph and make it appear to show something that never happened.

This may include:

• changing a photograph to make a child appear nude or partly nude;
• placing a child’s face into a false sexual image or video;
• creating completely artificial images that appear to show real children;
• producing cartoons or animations containing sexual representations of children.

These images are often described as deepfakes or AI-generated child sexual abuse material. The original photograph may have been completely innocent. It could have come from a social media profile, a school website, a sports club, a family event or a messaging application. The problem is therefore not only what children share themselves. Adults, schools, organisations, relatives and friends may also publish photographs that can later be copied and misused.

Synthetic Does Not Mean Harmless

Some people may believe that an AI-generated image causes less harm because the event shown in the image did not actually take place. This understanding is misleading. When the image represents a real child, the child may experience fear, humiliation, anxiety and a loss of control. The image can be copied, shared or used for blackmail. The child may also worry that friends, classmates or family members will see it and believe it is real. Even fully artificial images can contribute to a wider culture that sexualises children and treats child abuse as acceptable or normal. The guide also makes clear that, under UK law, indecent images of children may be illegal even when they have been created or altered using AI. Legal definitions and reporting procedures may differ between countries, but the central safeguarding principle remains the same: technology does not remove the harm or the need for an appropriate response.

Image Consent Must Become Part of Everyday Safety

The new guidance places strong emphasis on image consent. Image consent means that children should have a voice in decisions about whether their photographs are taken, published or shared. This is especially important as children grow older and become more aware of their identity and digital presence. Parents and carers can begin by reviewing the images they have already posted online. Useful questions include:

• Is the child’s face clearly visible?
• Can the school, location or daily routine be identified?
• Who is able to view or download the image?
• Is the child comfortable with the image being online?
• Is it still necessary for the photograph to remain public?

Families can also check whether schools, clubs, relatives or friends have published photographs of the child. Asking someone to remove a photograph is not an overreaction. It is a reasonable step when a parent or child no longer feels comfortable with an image being publicly available.

Parents Do Not Need to Understand Every Technology

Many parents feel that children know more about digital technology than they do. This can make conversations about AI feel difficult. However, adults do not need to be technical experts. The most important step is to create a relationship in which a child feels safe enough to speak when something goes wrong. Children should know that they will be listened to, believed and supported rather than blamed. A conversation can begin with simple questions:

“What AI tools have you used?”

“Have you seen images online that looked real but were actually fake?”

“What would you do if someone changed a photograph of you or a friend?”

“Who could you speak to if something online made you feel uncomfortable?”

These questions allow adults to understand what children already know and to discuss safety without creating panic.

What Should an Adult Do When a Child Asks for Help?

The first response can make a major difference. When a child says that an image has been created, altered or shared, adults should remain calm. The child should be told clearly that speaking about it was the right decision and that the situation is not their fault. Adults should avoid asking the child to repeatedly view or display the image. Relevant messages, threats and account information should be preserved because they may be needed as evidence. Where blackmail is involved, the child should stop communicating with the person responsible. Money should not be paid, as payment does not guarantee that the material will be deleted. The incident should be reported to the relevant platform and to the appropriate national law-enforcement or child-protection authority. Psychological and emotional support may also be needed, particularly when the child is experiencing shame, fear or social pressure. The guidance reminds families that support and removal services exist and that children should not have to manage the situation alone. 

A Challenge That Requires More Than Technical Solutions

AI-generated sexual abuse material is not only an online safety issue. It is connected to child protection, mental health, education, policing, victim support, legislation, platform responsibility and children’s rights. It also raises difficult questions for professionals:

How can schools protect photographs of pupils?

How should professionals speak to children about manipulated images?

What support is needed after an image has been shared?

How should platforms respond when harmful content can be created and distributed within minutes?

How can laws and safeguarding systems keep pace with rapidly changing technology?

These questions cannot be answered by one sector alone.

Keeping Child Protection in Step with Change

The Connect to Protect – Children & Youth at Risk International Conference, taking place in Athens on 5–6 October 2026, is designed to bring such contemporary challenges into the centre of professional and policy discussion. The conference will examine how risks affecting children are changing across both physical and digital environments. Online exploitation, digital manipulation, harmful content, grooming, technology-facilitated violence, mental health and the responsibilities of adults and institutions form part of a wider conversation about modern child protection. By engaging with new evidence and emerging threats, Connect to Protect aims to connect research, professional practice and policymaking. Being at the forefront of child protection does not mean reacting only after a new form of harm becomes widespread. It means identifying changes early, preparing professionals, informing families and ensuring that children’s rights remain central as technology develops. Artificial intelligence will continue to evolve. Child protection systems must evolve with it. The task is not to create fear around technology, but to ensure that innovation is accompanied by responsibility, informed adults, accountable institutions and effective support for every child who may be affected.

This article was informed by the June 2026 guidance “Child Sexual Abuse Material Generated by Artificial Intelligence: An Essential Guide for Parents & Carers”, developed by CEOP Education, the UK National Crime Agency and the Internet Watch Foundation.