AI, Deepfakes, Sextortion and ALN: What Welsh Parents Need to Understand About Emerging Online Risks

This is the third article in our online safety mini-series for ALN families. It follows Online Safety, ALN and the Children’s Wellbeing and Schools Act: Why Wales Needs Smarter Safeguarding and Assistive Technology, Parental Controls and ALN: Why Online Safety Advice Does Not Always Fit Disabled Children.

AI has changed the online safety conversation.

Some risks are not completely new. Grooming, bullying, coercion, blackmail, image-based abuse, fake friendship, harmful content and exploitation all existed before generative AI became widely available.

But AI can make some risks faster, cheaper, more convincing and harder to recognise.

A child or young person may now face risks involving fake images, fake voices, fake profiles, persuasive chatbots, manipulated screenshots, AI-generated sexual images, deepfakes, scams, blackmail or threats to share content that may not even be real.

For ALN families, this matters.

Some children and teenagers may be technically skilled but socially vulnerable. Some may rely heavily on online spaces because school has broken down, friendships are difficult, communication is easier online, or digital access is part of EOTAS, home education or learning.

The key message is simple:

A fake image can still cause real harm.

And the first safeguarding message should be even simpler:

Tell a safe adult quickly.

This article is not here to frighten families

This article is not about panic.

It is not about treating every child’s online life as dangerous.

It is not about blaming children.

It is not about blaming parents.

It is about naming risks clearly enough that families can act early.

A child should never be left believing they are to blame because someone else has used technology to exploit, threaten or shame them.

That is especially important where a child or young person has ALN.

Some children may not understand what has happened.

Some may be terrified that they will be punished.

Some may believe a fake image is their fault.

Some may think a threat must be obeyed.

Some may not know the difference between a joke, a dare, a scam, coercion, grooming or a criminal act.

Some may not be able to explain the situation clearly.

The first message needs to be:

You are not in trouble for asking for help. Tell a safe adult quickly.

What do we mean by deepfakes, sextortion and AI-generated abuse?

A deepfake is synthetic or manipulated media that can make it look or sound as if someone said or did something they did not say or do.

Sextortion is a form of blackmail where someone threatens to share sexual, intimate or embarrassing images, real or fake, unless the child does what they ask. That demand may be for more images, money, secrecy, contact, control or compliance.

In online safety conversations, parents may hear about:

• AI-generated sexual images;
• “nudified” images;
• fake intimate images;
• manipulated school photos;
• fake screenshots;
• cloned voices;
• fake profiles;
• AI chatbots pretending to be people;
• image-based blackmail;
• sextortion;
• threats to share real or fake images.

The important point is that the harm does not depend only on whether the image is “real”.

A fake image can still humiliate, frighten, isolate, coerce or damage a child.

A fake image can still be used to threaten a child.

A fake image can still be shared by others.

A fake image can still lead to bullying, distress, shame or self-harm risk.

A fake image involving a child may also raise serious legal and safeguarding concerns.

The first practical message: do not face the threat alone

If a child or young person is being threatened, the first steps are usually simple but urgent:

• do not pay;
• do not send more images;
• do not continue the conversation alone;
• tell a safe adult quickly;
• keep evidence where it is safe and lawful to do so;
• report through the appropriate safeguarding, police, school, platform or specialist route.

The child may feel ashamed.

They may feel frightened.

They may worry that they will be punished.

They may believe that doing what the person demands will make the problem go away.

But blackmail often escalates when the person threatening the child gets what they ask for.

The message to children needs to be clear:

If someone is threatening you, do not handle it alone. Tell a safe adult quickly.

What does the law already say?

The law is not empty, but it is still catching up with the speed and scale of AI-enabled risk.

The Online Safety Act 2023 places duties on online services and platforms. GOV.UK explains that the Act requires services to reduce the risk that their platforms are used for illegal activity, take down illegal content, stop children from accessing harmful and age-inappropriate content, and provide clear ways to report problems online [1].

Ofcom is the regulator for the Online Safety Act. Ofcom’s child safety guidance focuses on duties for services likely to be accessed by children, including risk assessments, safer design, age assurance and measures to protect children from harmful content [2], [3].

The Crown Prosecution Service guidance on indecent and prohibited images of children explains that computer-generated or AI-generated images can fall within the law where they appear to be photographs. The guidance says the law applies regardless of the method of production [4].

Victim Support explains that image-based sexual abuse can include threats to share intimate images, and that threatening to share an intimate image may be an offence whether or not the image exists [5].

So a child or young person should not assume that “it is fake” means “nothing can be done”.

At the same time, law often acts after harm has already happened.

Families still need prevention, accessible explanations, quick reporting routes and trusted adults.

Why this matters to the Children’s Wellbeing and Schools Act conversation

This article sits alongside our Children’s Wellbeing and Schools Act series because the same safeguarding question keeps appearing:

Does visibility actually lead to prevention?

The Act’s Wales-facing children not in school framework may make children outside school more visible to local authorities. Welsh Government has said that the children not in school measures include children who are electively home educated, children in flexi-schooling arrangements, and children receiving local authority EOTAS provision [6].

That may matter.

But visibility by itself will not teach a child what to do if someone uses AI to threaten, shame or exploit them.

A register will not tell a child:

• what sextortion is;
• why a fake image can still be reported;
• why they should not pay;
• why they should not send more;
• why shame helps the exploiter;
• who their safe adult is;
• how to ask for help if they cannot explain things easily.

Children outside school may need online safety education just as much as children in school.

Some may need it more, because online spaces may be where their learning, friendships, identity or communication now happen.

Why AI changes the practical risk

AI can change risk in several ways.

It can make fake content look more believable

A child may be shown an image, voice note, screenshot or message that looks convincing.

They may not know whether it is real.

They may panic before asking for help.

They may believe everyone else will believe it.

That fear can be enough for coercion.

It can make abuse easier to create

Some AI tools can create or manipulate images quickly.

That can lower the barrier for people who want to bully, humiliate, threaten or exploit others.

It can also make it easier for children themselves to experiment without understanding the consequences.

It can make blackmail more confusing

A child may be threatened with a real image, a fake image, or a claim that an image exists.

To the child, the difference may not matter in the moment.

They may feel trapped.

They may comply with further demands because they are frightened.

It can make fake intimacy more persuasive

AI chatbots and automated accounts can produce warm, responsive, personalised messages.

A vulnerable child or teenager may experience this as friendship, romance, support or emotional safety.

We have already written about when online engagement becomes emotional dependency, because attachment, algorithms and emotional pull can all matter when young people spend more time online [13].

It can move faster than adults can follow

Parents, schools, tutors and professionals may not know which tools are being used or what those tools can now do.

That is not because adults are careless.

It is because the technology is moving quickly.

This is why families and professionals need ongoing learning, not one-off online safety advice.

Why ALN children and teenagers may be more vulnerable

ALN does not mean a child cannot be safe online.

It does not mean all ALN children are naive.

It does not mean parents should panic.

But some children and young people may have specific vulnerabilities that online safety advice needs to understand.

An ALN child or teenager may:

• take words literally;
• trust people quickly;
• miss hidden motives;
• find it hard to read social cues;
• struggle to detect sarcasm, manipulation or coercion;
• be highly impulsive;
• seek belonging intensely;
• feel isolated from peers;
• rely on online friendship;
• have a strong special interest that draws them into specific online spaces;
• struggle to explain what has happened;
• freeze when frightened;
• feel intense shame;
• be more vulnerable to blackmail;
• not understand permanence, screenshots or sharing;
• be technically skilled but socially vulnerable.

That last point matters a lot.

A young person may know how to use AI tools, gaming servers, editing apps, private messages, multiple accounts or online platforms.

That does not mean they understand coercion, grooming, sexual pressure, financial scams, reputational harm or the emotional consequences of sharing information.

Digital skill is not the same as social safety.

Children outside school may need extra thought

Children outside school may spend more time online.

That can include children who are:

• electively home educated;
• receiving EOTAS;
• on online tutoring packages;
• unable to attend school because of anxiety, burnout or health needs;
• recovering from school trauma or placement breakdown;
• socially isolated;
• using online communities for friendship;
• using technology for communication or assistive access.

Online access may be a lifeline.

It may provide education, friendship, identity, creativity, coding, gaming, art, learning or community.

But it can also increase exposure to risk if the child is isolated, lonely, dysregulated, seeking belonging or communicating mainly through private online spaces.

That is why monitoring whether a child is “not in school” does not automatically answer the online safety question.

A register does not tell you whether a child understands AI blackmail.

It does not tell you whether an online tutor platform has safe messaging rules.

It does not tell you whether a child is being drawn into unsafe communities.

It does not tell you whether a young person understands that a fake image can still be reported.

It does not tell you whether the child knows which adult to tell.

This is why LWL keeps saying that visibility is not the same as prevention.

What recent warnings show

Recent UK reporting has highlighted concern that identifiable school photos can be misused to create AI-generated sexual images for blackmail. The Guardian reported that experts and authorities were urging schools to remove or limit identifiable pupil photographs online after criminals used AI to create sexually explicit images from ordinary school images [10].

That does not mean every school photo will be misused.

It does mean parents, schools and organisations should think carefully about public images of children, especially where names, school uniforms, locations or other identifying details are visible.

It also means families need to understand that the image used to threaten a child may not have come from a private message.

It might have been taken from somewhere ordinary:

• a school website;
• a club page;
• a public social media post;
• a group photo;
• a competition announcement;
• a sports page;
• a fundraising page;
• a family post;
• a profile picture.

For ALN families, the practical question is not only “what has my child shared?”

It may also be:

What images of my child are already publicly available, and do they need to be there?

What children need to hear

Children and young people need clear, simple messages.

These messages should be repeated calmly and often, not only after something has gone wrong.

Useful messages include:

• If someone asks you to keep a risky secret, tell a safe adult.
• If someone threatens to share an image, tell a safe adult quickly.
• If someone says they have a picture of you, tell a safe adult quickly.
• If an image is fake, still tell someone.
• If you shared something and now feel scared, still tell someone.
• If someone asks for money, pictures, passwords or more images, tell someone.
• If you are worried you will be in trouble, tell someone anyway.
• You are not responsible for someone else choosing to threaten or exploit you.
• A fake image can still be harmful, and you deserve help.
• Asking for help early is safer than trying to handle blackmail alone.

For some ALN children, these messages may need to be visual, concrete, repeated, scripted, practised or built into social stories, communication systems, IDP support or trusted adult routines.

The first safeguarding message should always be:

Tell a safe adult quickly.

What parents can do without panicking

Parents cannot prevent every risk.

But they can reduce some risks and improve the chance that a child asks for help early.

Practical steps may include:

• checking what public images of the child exist online;
• avoiding public posts that combine face, name, school, location and routine;
• asking schools, clubs or groups how images are used;
• reviewing image consent forms;
• avoiding public captions that identify children unnecessarily;
• checking privacy settings on family social media accounts;
• talking calmly about fake images and threats before anything happens;
• agreeing who the child can tell if they are scared;
• agreeing that the child will not be punished for asking for help;
• checking whether online tutoring or EOTAS platforms allow private messaging;
• asking whether chats, recordings or contact routes are monitored;
• helping the child understand screenshots, forwarding and permanence;
• making sure the child knows not to pay, send more images, or continue a threatening conversation alone;
• keeping evidence if a threat happens, without forwarding harmful images unnecessarily;
• seeking urgent help through the appropriate safeguarding, police, school, platform or specialist route.

Parents should not have to become digital forensic experts.

They need practical, accessible routes to help.

What to preserve as evidence

Parents may feel an understandable urge to delete everything immediately.

But where threats, grooming, blackmail or image-based abuse may be involved, it may be important to preserve messages, usernames, profile links, account names, platform details, timestamps, links or screenshots where it is safe and lawful to do so.

Do not forward or share illegal images.

Do not post the material publicly.

Do not confront the person threatening the child in a way that increases risk.

Seek appropriate safeguarding, police, school, platform or specialist advice.

The aim is not to investigate everything yourself.

The aim is to avoid accidentally destroying information that the appropriate authority or platform may need.

What not to do if a child discloses

If a child says they are being threatened, blackmailed or shamed online, the first adult response matters.

Try not to start with:

• “Why did you do that?”
• “How could you be so stupid?”
• “I told you this would happen.”
• “Give me your phone, you’re never going online again.”
• “This is your own fault.”

Those reactions may make a child stop talking.

Instead, the first response needs to be calm enough to keep the door open.

A better starting point is:

• “Thank you for telling me.”
• “You are not alone.”
• “We will get help.”
• “Do not send anything else.”
• “Do not pay anyone.”
• “We need to save the evidence safely.”
• “You are not to blame for someone threatening you.”
• “A fake image can still be reported.”
• “We will take this seriously.”

This does not mean ignoring what happened.

It means responding in a way that keeps the child safe enough to keep talking.

Where to report or seek help

The right route depends on what has happened.

If there is immediate danger, a child is at risk, or a crime may be taking place, families should use the appropriate emergency, police, safeguarding or local authority route.

If the issue involves a school, EOTAS provider, tutor, club or organised activity, parents may also need to contact the designated safeguarding lead or responsible safeguarding person.

If the content involves child sexual abuse imagery online, the Internet Watch Foundation provides a UK route for reporting suspected child sexual abuse images or videos online [9].

If a young person is worried about an intimate image being shared online, they may also need specialist advice about reporting, takedown routes and support. Victim Support has information on image-based sexual abuse and explains that threats to share intimate images can be an offence [5].

The most important point is this:

Do not let shame, confusion or fear delay asking for help.

What schools, local authorities and providers should understand

Schools, local authorities, EOTAS providers, online tutoring providers and youth organisations need to understand that AI risk is now part of safeguarding.

That does not mean children should be frightened away from technology.

It means adults should understand the risks well enough to design safer systems.

They should consider:

• whether pupil photos need to be public;
• whether names, uniforms, schools or locations are visible;
• whether image consent forms are specific enough;
• whether online tutoring platforms allow private messaging;
• whether chat, video and recording settings are safe;
• whether children know how to report AI-generated abuse;
• whether ALN children are taught using accessible language;
• whether online safety lessons are concrete enough;
• whether parents understand what tools are being used;
• whether children outside school are missing online safety education;
• whether IDP, EOTAS or transition planning should include digital safety.

For ALN children, online safety teaching may need more than one assembly, one worksheet or one generic internet safety lesson.

It may need repetition, visual support, concrete examples, trusted adults and practice.

Where LWL’s parent community can help

Parent peer support cannot replace safeguarding advice, police, social care, school safeguarding routes, legal advice, health services or specialist technical support.

But trusted parent communities can help families act earlier.

LWL’s parent peer community has been useful in practical safeguarding terms. Over the years, parents have sometimes used the community to sense-check concerns, ask where to go next, or understand whether something may need urgent action. Where parents have identified online risks that sit beyond peer support, the community has helped signpost them promptly towards the appropriate authorities or specialist routes. On occasion, that has included the police.

We do not discuss individual cases publicly.

We do not encourage parents to ignore formal safeguarding routes.

But the experience matters.

When a parent is unsure whether something has crossed a line, a trusted community can help them stop minimising the concern and move towards the right route.

That is part of prevention.

It is also why open, honest, non-shaming discussion matters.

Parents and professionals are trying to keep up with technology that changes faster than many formal training routes can update. Peer communities cannot replace training, but they can help identify emerging questions quickly.

Where LWL fits

LWL’s wider approach is Prevent, Bridge, Progress.

Online safety belongs in all three.

Prevent means helping families understand risks earlier, before harm escalates.

Bridge means supporting families in the messy middle, where children may already be outside school, online more often, socially isolated, using assistive technology or recovering from educational breakdown.

Progress means helping children and young people move towards safer independence, better digital understanding, education, community, skills and future routes.

This is not about making parents frightened of technology.

It is about helping families, professionals and policy-makers ask better questions.

Because the goal is not to cut ALN children off from digital life.

The goal is to help them access digital life more safely.

The key message

AI changes the online safety conversation.

A fake image can still cause real harm.

A threat can be frightening even when the image is not real.

A child can be digitally skilled and socially vulnerable.

A child outside school may rely heavily on online spaces for learning, friendship and identity.

A child should never be left believing they are to blame because someone else has used technology to exploit, threaten or shame them.

The first safeguarding message should always be:

Tell a safe adult quickly.

For Wales, the challenge is to build online safety that is realistic, ALN-aware, disability-aware and practical.

Not panic.

Not blame.

Not blanket restriction.

Not silence.

Children need protection, access, trust and routes to help.

Families need guidance that understands the world their children are actually navigating.

Legal disclaimer

This article is for general information only. It is not legal advice, safeguarding advice, criminal justice advice, technical advice or clinical advice.

Families facing immediate online risk, grooming, exploitation, threats, image-based abuse, self-harm risk, criminal behaviour or safeguarding concerns should seek urgent support from the appropriate emergency, safeguarding, policing, school, local authority, health or specialist advice route.

References

[1] GOV.UK, “Online Safety Act: explainer,” updated 24 Apr. 2025. Available: https://www.gov.uk/government/publications/online-safety-act-explainer/online-safety-act-explainer

[2] Ofcom, “Protecting children from harms online,” accessed May 2026. Available: https://www.ofcom.org.uk/online-safety/protecting-children/protecting-children-from-harms-online/

[3] Ofcom, “Children’s Safety Codes and Guidance,” accessed May 2026. Available: https://www.ofcom.org.uk/online-safety/protecting-children/childrens-safety-codes-and-guidance/

[4] Crown Prosecution Service, “Indecent and Prohibited Images of Children,” accessed May 2026. Available: https://www.cps.gov.uk/prosecution-guidance/indecent-and-prohibited-images-children

[5] Victim Support, “Image-based sexual abuse,” accessed May 2026. Available: https://www.victimsupport.org.uk/crime-info/types-crime/image-based-sexual-abuse/

[6] Welsh Government, “Children’s Wellbeing and Schools Bill and elective home education,” 26 Feb. 2026. Available: https://www.gov.wales/childrens-wellbeing-and-schools-bill-and-elective-home-education-html

[7] Welsh Government, “Additional learning needs (ALN): decision-making and communication,” 27 Mar. 2026. Available: https://www.gov.wales/additional-learning-needs-aln-decision-making-and-communication-html

[8] Welsh Government, “Additional learning needs (ALN) and education otherwise than at school (EOTAS),” 27 Mar. 2026. Available: https://www.gov.wales/additional-learning-needs-aln-and-education-otherwise-school-eotas-html

[9] Internet Watch Foundation, “Report suspected child sexual abuse images or videos,” accessed May 2026. Available: https://report.iwf.org.uk/

[10] The Guardian, “UK schools should remove pupils’ online photos as AI blackmail threat grows, say experts,” 8 May 2026. Available: https://www.theguardian.com/technology/2026/may/08/uk-schools-remove-pupils-photos-online-ai-blackmail-threat-grows

[11] Learn Without Limits CIC, “Online Safety, ALN and the Children’s Wellbeing and Schools Act: Why Wales Needs Smarter Safeguarding,” May 2026. Available: https://blog.learnwithoutlimitscic.org/online-safety-aln-childrens-wellbeing-schools-act-wales/

[12] Learn Without Limits CIC, “Assistive Technology, Parental Controls and ALN: Why Online Safety Advice Does Not Always Fit Disabled Children,” May 2026. Available: https://blog.learnwithoutlimitscic.org/assistive-technology-parental-controls-aln-online-safety/

[13] Learn Without Limits CIC, “When ‘Engagement’ Becomes Emotional Dependency: Why ALN Children Are Disproportionately Exposed,” Feb. 2026. Available: https://blog.learnwithoutlimitscic.org/2026/02/when-engagement-becomes-emotional.html