Why photos now need a stronger safeguarding standard
AI abuse in schools has changed the safeguarding test for images of children. This does not mean every photograph should be removed. It means schools, charities and community organisations must ask harder questions before publishing, storing or sharing images. Consent, dignity, necessity and review now matter more than ever.
Schools, charities and community organisations use real photographs for good reasons. Images can show children learning, young people leading, volunteers helping, patients being supported and communities coming together. They can build trust, show impact and give dignity to people whose stories deserve to be seen.
That must not be dismissed.
The issue is not whether images should ever be used. The issue is whether organisations now use them carefully enough for the AI age.
The UK Safer Internet Centre has briefed schools on image safety after concerns about AI blackmailers using pupil images from websites and social media. Its advice focuses on managing children’s image security, not banning all photographs. It asks schools to think more carefully about how images are shared, protected and reviewed. That is the right balance. Visibility matters. Safeguarding matters too.
For years, schools have used photographs to celebrate achievement. A sports day picture. A class project. A student award. A school trip. Most of that use has been positive and well intentioned.
However, good intent is no longer enough.
AI tools have changed what can happen to an image after it leaves an organisation’s control. A clear photograph can be copied, manipulated and misused. A child’s name, uniform or school identity can add further risk. A public image can become material for humiliation, coercion or extortion.
That does not mean every image is unsafe. It means the risk assessment has changed.
The old question was:
Do we have permission to use this photograph?
The new question must be:
Is this image necessary, proportionate, consented to, reviewed and safe enough to publish in the current digital environment?
That is a higher standard, but it is not an impossible one.
Responsible image use should not disappear from schools, charities or public-interest work. Children and young people should not become invisible because criminals abuse technology. Survivors, patients, volunteers and campaigners should not be erased from public life because risk exists.
The task is to reduce avoidable risk.
That means choosing images with care. It means avoiding unnecessary full names. It means checking whether a child’s school, location, routine or vulnerability is too visible. It means reviewing old galleries and removing images that no longer need to remain online. It also means giving parents, carers, children and adults a clear route to ask for an image to be removed.
Government guidance recognises that schools may use photos and videos for educational activities, marketing, newsletters, social media, school trips, sports events, performances and classroom displays. It also says identifiable images count as personal data and must be handled in line with data protection legislation.
That distinction is vital. Schools do not need a blanket ban. They need better judgement.
Many organisations rely on image consent forms. Those forms still matter. However, some were written before AI misuse became a mainstream safeguarding concern.
A parent may have agreed to a child’s photograph appearing in a newsletter. A student may have agreed to an image being used to celebrate an award. A volunteer may have agreed to appear in a campaign photo.
That does not mean they understood how easily an image could later be copied, scraped, altered or misused.
Consent should be clear, current and meaningful.
The Information Commissioner’s Office says data protection law is likely to apply where photos or videos are taken for official school use, such as promotional material, and pupils or parents must be informed how the images will be used.
In practical terms, organisations should ask:
Does the consent wording explain where images may appear?
Does it explain social media use?
Does it say how long images may remain online?
Can consent be withdrawn?
Does the person understand whether their name will appear?
Does the organisation have a process for removing images?
Does the policy now mention AI manipulation, image misuse or sextortion?
These questions are not bureaucratic. They are part of modern safeguarding.
At Freedom Charity, we have always believed that education can protect children before harm becomes irreversible. Freedom’s safeguarding work includes school resources, professional awareness, But It’s Not Fair, Cut Flowers and PSHE-accredited forced marriage and FGM lesson plans. Freedom’s PSHE resources were produced with the PSHE Association and support teachers to approach forced marriage and FGM in a sensitive, age-appropriate and safeguarding-led way.
Those resources exist because children need information, visibility and safe routes to help.
That is why this issue must be handled carefully. We should not respond to AI abuse by making children invisible. Silence has never protected children from abuse.
However, public-facing images must now sit within a stronger safeguarding framework.
Schools and charities can still use images. They should do so with consent, dignity and review.
The phrase “online safety” can sound too soft. AI abuse in schools can involve child protection, image-based abuse, sexual coercion, extortion, reputational harm and violence against women and girls.
The Internet Watch Foundation’s 2025 Annual Data and Insights Report recorded 451,210 reports assessed and 311,610 reports confirmed to contain or lead to child sexual abuse material. Its report also identifies generative AI and violence against women and girls as emerging and persistent harms.
A fake image can cause real harm. A manipulated image can create fear, shame and isolation. A child may worry that adults will blame them. A young person may be threatened with exposure. A family may not know where to turn.
The first response must be protection, not judgement.
AI abuse can affect boys, men, staff, volunteers and public figures. However, sexualised image abuse has a strong gendered dimension.
The Internet Watch Foundation reported that girls accounted for 97% of AI-generated child sexual abuse imagery cases it assessed in 2025. It also warned about the use of AI-generated images in sexual extortion schemes, including the ease with which nudifying tools can create compromising images.
This matters because shame is often used as a weapon against girls and women. A fake image can still be used to threaten someone’s education, reputation, family relationships or personal safety.
The image may be artificial. The coercion is not.
Professionals must understand that distinction.
A good image policy should not frighten organisations into silence. It should help them make better decisions.
Schools, charities and community organisations should review:
This is not about stripping websites of life. It is about using real people’s images with care, dignity and purpose.
Safer image use does not have to be complicated.
Organisations can use group images where individuals are less identifiable. They can take photographs from behind, from a distance or from an angle that still shows activity without exposing a child’s face. They can avoid full names. They can use first names only where appropriate. They can blur faces where needed. They can remove old images after a set period.
They can also avoid linking images to sensitive personal stories unless the person has given informed and specific consent.
Strong non-identifying images can work well. These may include hands working, classrooms from behind, campaign materials, books, displays, feet on a sports field, artwork, event banners, backs of heads, symbolic objects or carefully cropped scenes.
The point is not to hide people. The point is to avoid unnecessary exposure.
The UK Government has recognised that AI can be used to facilitate child sexual abuse, violence against women and girls, fraud and other crimes.
Section 138 of the Data (Use and Access) Act 2025 came into force on 6 February 2026. It created offences relating to creating, or requesting the creation of, a purported intimate image of an adult without consent or reasonable belief in consent.
For children, sexual images and pseudo-images raise serious child protection and criminal law issues. The Crime and Policing Act 2026 factsheet states that the Act introduces a new offence criminalising AI models optimised to create child sexual abuse material.
However, law alone will not protect children. Organisations need practical policies, trained staff and clear escalation routes.
Every school, charity and community organisation should carry out an image audit.
That does not mean deleting everything overnight. It means checking what is online, why it is there and whether it still needs to be public.
The first priority should be images that are:
Organisations should then update image policies, consent wording and staff guidance. They should also make sure someone has responsibility for removing images quickly if concerns arise.
A good question for every image is:
Does this image still serve a clear purpose, and can we use it in a safer way?
Parents and carers do not need to become technology experts. They can ask simple questions.
Where will this image appear?
Will my child’s name be used?
Will the image be shared on social media?
How long will it stay online?
Can I withdraw consent?
What happens if the image is copied or misused?
Does the organisation review old photographs?
Does the policy cover AI manipulation?
Adults who appear in charity, health or campaign images should be able to ask similar questions. Consent should never be treated as a one-off signature that lasts forever regardless of context.
If an image is misused, the person in the image is not responsible.
Not the child.
Not the parent.
Not the school that acted in good faith.
Not the charity that used a photograph responsibly.
Responsibility sits with the perpetrator and with any system that enables abuse to spread without effective prevention, reporting and removal.
That point matters because shame helps abusers. Clear safeguarding language helps victims.
Before publishing an image, every organisation should ask:
Would we still use this photograph if we understood how easily it could be copied, altered, sexualised or used to threaten the person in it?
If the answer raises concern, the image should be reviewed before publication. The organisation should consider whether the person needs to be identifiable, whether a name is necessary, whether the image should be time-limited and whether a safer version would achieve the same purpose.
This is not about fear. It is about judgement.
Children, students, patients, survivors, volunteers and campaigners should not be made invisible. Their stories matter. Their dignity matters. Their safety matters too.
AI abuse in schools has changed the safeguarding landscape. The response should be calm, practical and urgent: not a ban on images, but a higher standard for using them.
AI abuse in schools includes the use of artificial intelligence tools to alter, sexualise, impersonate or misuse images, videos or voices connected to pupils, staff or the school community.
No. Schools should take a proportionate approach. They should review images, avoid unnecessary identification, update consent processes and remove images that create avoidable risk.
Yes, where image use is necessary, consented to, dignified and reviewed. Charities should avoid unnecessary names, sensitive contexts and images that expose vulnerable people to avoidable risk.
Consent matters, but it must be informed and current. People should know where images may appear, how long they may remain online, whether names will be used and how they can ask for removal.
Yes. A fake image can still cause fear, shame, bullying, coercion and reputational harm. The image may be artificial, but the impact can be real.
Parents should ask where images are published, whether names are used, how long images remain online, whether consent can be withdrawn and what the school will do if an image is misused.
Aneeta Prem MBE is a UK author, human rights campaigner and safeguarding expert. She founded Freedom Charity and has worked for many years on forced marriage, dishonour abuse, FGM, coercion, child protection and the rights of women and girls. Her safeguarding books, But It’s Not Fair and Cut Flowers, are used to help young people and professionals recognise abuse and seek help safely.
UK safeguarding charity working to protect children, young people and survivors from forced marriage, FGM, dishonour abuse and coercion.
Link: https://freedomcharity.org.uk/
Education resources supporting schools to teach forced marriage and FGM safely, clearly and responsibly.
Link: https://freedomcharity.org.uk/pshe-accredited-forced-marriage-and-fgm-lesson-plans/
Aneeta Prem’s forced marriage safeguarding book for schools, professionals, families and young people.
Link: https://freedomcharity.org.uk/but-its-not-fair-forced-marriage-book/
Freedom Charity uses the term dishonour abuse because there is no honour in coercion, control or violence.
Link: https://freedomcharity.org.uk/dishonour-abuse/
A safeguarding article explaining why dishonour abuse must be recognised as child abuse, coercive control and a serious protection concern.
Link: https://freedomcharity.org.uk/dishonour-abuse-is-child-abuse/
Advice for schools on image safety and the risk of AI blackmail using pupil photographs.
Link: https://saferinternet.org.uk/blog/schools-given-advice-on-image-safety-to-keep-ahead-of-threat-from-ai-blackmailers
Official UK Government guidance on taking, storing and sharing photos and videos safely in schools.
Link: https://www.gov.uk/guidance/data-protection-in-schools/taking-and-using-photos-and-videos-and-using-cctv-in-schools
Official data protection guidance explaining when school photographs are likely to fall under data protection law.
Link: https://ico.org.uk/for-the-public/schools/photos/
Evidence on child protection, online image-based harm and child sexual abuse material reports.
Link: https://www.iwf.org.uk/annual-data-insights-report-2025/
Evidence on AI-generated child abuse imagery, synthetic image harm and gendered risk.
Link: https://www.iwf.org.uk/annual-data-insights-report-2025/image-insights/ai-generated-images/
Official source confirming commencement of section 138 on purported intimate image offences.
Link: https://www.gov.uk/guidance/data-use-and-access-act-2025-plans-for-commencement
Current legal source on offences relating to AI models optimised to create child sexual abuse material.
Link: https://www.gov.uk/government/publications/crime-and-policing-act-2026-factsheets/crime-and-policing-act-2026-child-sexual-abuse-material-factsheet
By Aneeta Prem MBE
Published: 9 May 2026
Updated: 26 May 2026
The murder of Henry Nowak raises difficult questions about policing, evidence and the duty to preserve life
The death of Twisha Sharma has reopened a global question about dowry abuse, coercive control, family reputation and why women are still judged by what their families can provide.
Afghanistan's child marriage is worsening as hunger, aid cuts and Taliban family law place girls at risk. Aneeta Prem writes on why silence is not consent.
AI abuse in schools is not a reason to remove every photograph. It is a reason to use images with stronger consent, clearer purpose, and better safeguarding judgment.
Laws can punish abuse after it happens. Preventing dishonour abuse requires something harder: educating children early, supporting victims safely, and challenging the beliefs that allow coercion, fear and control to thrive behind closed doors.
A Child Cannot Consent to Marriage. The Nottingham child marriage case shows why UK law had to change. Since February 2023, under-18 marriage has been a criminal offence in England and Wales, including some overseas and religious ceremonies, even where coercion is not proved.