The Apps That Undress Learners: A Crisis Schools Cannot Ignore
Schools are left to punish children for harm that app stores, payment providers and platforms made cheap, easy and profitable.
Sign up for my Substack daily AI newsletter here.
Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/
A learner’s photo, taken for a sports team page or a prize-giving post, can be turned into a fake explicit image in less than a minute by a classmate with a phone and no technical skill. Most of the targets are girls. The harm is real even when everyone knows the picture is fake, and once it has travelled through a class group chat, no apology can recall it. Schools are usually the first institution to hear about it and the least equipped to deal with it. This article argues that our response has been aimed at the wrong end of the problem. Before a school expels a 14-year-old, someone should ask who built, rated, advertised and profited from the tool he used.
CONTEXT AND BACKGROUND
A WIRED and Indicator analysis found nearly 90 schools and more than 600 students affected by AI-generated nude images since 2023, across at least 28 countries, and noted that schools routinely publish pupils’ photographs online. UNICEF reports that at least 1.2 million children in 11 countries disclosed that their images had been manipulated into sexually explicit deepfakes in a single year.
At a recent Film and Publication Board webinar, experts said these apps are rampant in schools and are often used by boys to target girls they like. Section 16 of the Cybercrimes Act makes the non-consensual disclosure of intimate images an offence, including simulated images, with up to 3 years’ imprisonment. The relevant Films and Publications Act offence, however, is limited to photographs and to intent to cause harm, and no law bans the development or distribution of nudifying apps themselves. We punish the image, not the tool.
INSIGHT AND ANALYSIS
The Tech Transparency Project found that Apple’s and Google’s own search results, advertisements, and autocomplete suggestions steered users towards nudify apps. The apps it identified had been downloaded 483 million times and earned more than $122 million in lifetime revenue, and 31 of them were rated suitable for minors. Apple removed 15, and Google removed 7, only after the researchers and Bloomberg News shared their findings. Removal after a journalist asks is not a safety system.
Other jurisdictions have moved upstream. Britain’s Crime and Policing Act 2026 creates an offence of making or supplying tools designed to generate intimate images, with up to 3 years in prison, a defence for tools with consensual uses, and a duty on platforms to remove content offering them.
Ofcom now expects platforms to use hash matching, which recognises copies of known images, and can fine firms up to 10% of global revenue. In the United States, platforms must remove such images within 48 hours of a valid request.
I have previously written about this, arguing that schools should treat deepfake pornography as a severe form of cyberbullying, supported by consent education and counselling. That remains necessary, because schools still shape peer culture, yet it places most of the burden on the institution with the least power over the tools. Elsewhere in Africa, Nigeria’s technology regulator has warned that AI-generated explicit images are fuelling sextortion and that children are especially vulnerable.
IMPLICATIONS
The strongest counter-argument is that bans push these tools offshore, onto open-source software or into encrypted channels beyond any domestic regulator, and that loosely drafted laws could catch satire or ordinary photo editing. A ban will not end the problem, but it removes app store reach, advertising and payment channels, which is where scale comes from, and the British defence for consensual uses shows that overreach can be designed out. Many perpetrators are also children. Prosecuting a 13-year-old rarely repairs anything, so most school cases call for restorative and educational responses, with criminal routes reserved for threats, extortion and adult offenders.
Prevention therefore needs duties at every level. Government should extend the law to the supply of these tools and ask app stores to account for their age ratings. Boards of banks, payment providers, telecoms operators and advertising platforms should treat nudify services as prohibited merchants and customers. School governing bodies should name AI-generated sexual images in their codes of conduct, adopt a plan for the first 72 hours that contains the spread, protects evidence and puts the affected learner’s wellbeing first, and review how learners’ photos are published.
CLOSING TAKEAWAY
The crisis in our schools is real, yet it was manufactured elsewhere. The tools used to fake images of learners pass through a supply chain of developers, app stores, advertisers and payment providers before they reach a classroom, and each link in that chain can be made to carry responsibility. Schools must still prepare, teach consent and stand by the children who are harmed. Prevention will only work, however, when the companies that profit from these tools are asked to answer first. Our children deserve protection that starts long before the photo is taken, not discipline that arrives after the damage is done.
Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net
Comments