The implementation of age verification (AV) technology has long been waiting in the wings but finally came into effect 25th July 2025. It had been lauded as the corner stone provision of the Online Safety Act 2023, whose stated aim is to protect children online. On the surface, AV seems like a common-sense policy β one that few would argue against in its ambitions. We all understand the premise that certain items which are unsuitable should be held out of the reach of children. We all remember being IDed for alcohol or cigarettes back in the day. And yet, within a matter of days of the policy coming into practice, a petition to repeal the Act garnered over half a million signatures. The petition has been only the tip of the iceberg in backlash, which raises the question: how has a previously popular common-sense provision become the latest politicised culture war issue?
In this third instalment of our blog series –Β Beyond bans: rethinking phone-free childhoods and age verification through safe uncertainty,Β we unpack the current debate and concerns around the implementation of AV and what it means for young people, their rights and safety.Β
AV has been seen as the βsilver bulletβ solution in the quest of protecting children online and βmaking Britain the safest place for children to go onlineβ as David Cameron declared all the way back in 2013 at the NSPCC conference:
βI want to talk about the internet, the impact itβs having on the innocence of our children, how online pornography is corroding childhood and how, in the darkest corners of the internet, there are things going on that are a direct danger to our children and that must be stamped out. Now, Iβm not making this speech because I want to moralise or scaremonger but because I feel profoundly, as a politician and as a dad, that the time for action has come. This is, quite simply, about how we protect our children and their innocence.β
This speech set the tone for all the policy that has followed. Whilst the Online Safety Act aims to protect children from all harmful content β this has become synonymous with protecting children from pornography. This is a reasonable position. Indeed, before the days of the internet, the only place an adult could access pornography was through a licenced sex shop or adult cinema β the kind of places you would find in the run-down areas of town, where the shop would have blacked out windows and no child could enter. Teens would have to make do with discarded magazines in hedges or copies of the Daily Sport β or occasionally a friend would discover their parentβs hidden VHS collection.
Now the guard rails are off. No longer do teens need to search the hedgerows β porn has been only a click of a mouse away. The internet is awash with multiple free to view aggregate adult websites, whose only gatekeeping is a pop-up box with a warning βthis site contains adult sexual materialβ followed by the *intensely challenging*Β security safeguard of a click box asking if the visitor is over 18. In the history of the internet, no 15-year-old has been put off and clicked βNoβ β well-aware the sex police will not swoop in if they lie about their age.[1]
It is a reasonable ambition to want to move online pornography back to the proverbial top-shelf of the internet and out of the unsuspecting fingers of children who are too young to understand what they are watching. No child should accidentally stumble across adult sexual material online and there should be better safeguards in place. However, the criticism of the Online Safety Act is the safeguards now in place are not only disproportionate and an attack on civil liberties – but could actually make young peopleΒ less safeΒ online.Β Β The key is the dangerously vague wording of the Act:Β protecting children from online harms. This raises the question,Β whoΒ gets to decide what is harmful? And at what cost? For example, how does this infringe on young peopleβs rights to privacy (A16), access to information (A17) and participation under the United Nations Convention of the Rights of the Chid (UNCRC) Does the infringement of childrenβs privacy and rights not count as a harm?Β
Crucially, these questions extend far beyond pornography to a widening range of online content and activities now captured under the banner of βharmβ. When protection is defined so broadly, the risk is that we restrict not just pornography, but the very digital spaces where young people learn, connect and seek support, whilst undermining the relational communication that enables young people to talk openly with adults about what they encounter online.Β
Clamping down on digital childhoods:
The idea of AV is not new, and many of the criticisms which are currently being levelled at the policy should not come as any surprise to politicians as they have been raised time and time again. AV was part of the flagship policies included in The of the Digital Economy Bill (2017) and came close to being implemented back in October 2019, after already having been delayed. However, it was abandoned after immense opposition, and the realisation the policy in its current form was unworkable. Opposition mainly came from concerns surrounding privacy issues and security of personal information, but surprisingly there were also calls from many online safety experts who counterintuitively suggested AV may put young people at further risk (Blake, 2019)[2], none of which have been addressed in the interim.
Nevertheless, age verification is symptomatic of a broader trend in political discourse calling for stricter regulation of digital childhoods. There is a race between nations to enforce more stringent safeguards on childrenβs access to social media and mobile phones.Β Australia is proposing banning all teenagers from social media before the age of sixteen, in response to campaigns such asΒ Smartphone free childhood,Β andΒ Fight the new drug,Β whilst in response our own UK government is suggestingΒ time limiting young peopleβs access to social media to two hours a day, although it is unclear how this would be implemented. Whilst many parents may initially welcome such suggestions, we need to consider the broader implications for adolescent mental health and connection amongst peers. Like many popular policy proposals, the devil is in the detail β as in, there is none. There has been no consideration of the fact (whether we like it or not) teenagers build and maintain their relationships to a large part online and these polices would have detrimental effects for our young peopleβs wellbeing.Β
Yes, we all agree children should spend more time playing outside in the park, but the reality is somewhat different in practice. Many outside spaces have been purposefully designed to be hostile to young people. Successive governments have sold off playing fields, youth services have been gutted and groups of young people congregating in public spaces are seen as antisocial nuisance to the point many shops and spaces employ mosquito boxes that emit high frequency sounds that are designed to cause discomfort but are only audible to young peopleβs ears.
In reality, young people find solace, build connection and primarily seek information and advice via online platforms, message boards and social media. It is their first port of call when feeling anxious and is often a vital lifeline, especially for vulnerable and marginalised young people who may find it difficult or unsafe to speak to the adults in their lives β who themselves may pose a safeguarding risk. Yet AV focuses on restricting where young people go, rather than engaging with why they are drawn to these digital spaces for support, information and belonging.
To a large extent, many of these polices seem performative, rather than evidence-based in their conception, or effective in their application. I remember speaking to young people back in 2018-19 as the original deadline for the βporn blockβ β as AV was better known in those days β came closer to its threated implementation. Even back then young people were talking about Virtual Private Networks (VPNs) and other workarounds, whereas the adults seemed oblivious. And yet, whilst the policies may be performative, they can cause real harm as they are deeply flawed and misconceived. There are clear privacy risks with this new policy, and overreach as the definition of what is considered βharmfulβ content bleeds far past the bounds of pornography. Most importantly, AV can easily become a distraction as fund are diverted from proven methods of supporting healthy sexual development of adolescents and push young people further away from seeking help of safe adults, as online behaviours become further demonised[3].
In this manner, AV provisions represent a potential case of unsafe certainty: policies that attempt to ensure safety but risk undermining it through rigidity, false assurance and lack of reflexivity. Rather than closing down risk, such measures may push young people into more unregulated and unsafe digital spaces, where they are less likely to seek help. Worse, these policies may discourage open dialogue between adults and young people. If regulation is presented as sufficient or absolute, the perceived need for ongoing communication and relational trust may diminish. When harm occurs β as it inevitably will β young people may feel less able to disclose, fearing blame or further restriction.
Jimmy Savile sympathisers β government response to the backlash:
Unfortunately, there seems to be little compromise in the governmentβs response to raised concerns and backlash, which in itself seems odd, considering the Online Safety Act was not a Labour policy, but was passed by the previous Conservative government. AV has quickly become a culture war issue across the political spectrum. Whilst civil liberty groups and the likes of Nigel Farage are unlikely bedfellows, both seemingly to be on the same side of the argument, having condemned the policy. In response, MP Peter Kyle, the Secretary for Science, Innovation and Technology has vehemently defended the policy and when questioned about Nigel Farageβs vocal criticism, went so far as to claim anyone who disagrees with the policy βis on the side of Jimmy Savileβ¦β which would suggest they have little in way of argument to defend the policy, if they are resulting to name calling in attempt to shut down the conversation. Peter Kyle has since doubled down on his remarks and evoked the equivalent of the meme made famous by Helen Lovejoy in the Simpsonβs βWonβt someone please think of the children?β as he reiterated βverifying your age keeps children safe.β
What does AV look like in practice?
Under the Online Safety Act, it is platforms themselves who are given responsibility to protect children from harmful content by verifying the age of their users to ensure they have access only to age-appropriate content. If platforms fail in their duty, they face huge fines of up to Β£18 million or 10% of their global turnover, whichever is greater β which means the stakes are high β especially for smaller platforms.
As a result, many platforms are understandably airing on the side of caution, age gating their whole platforms or bringing in blanket bans for topics which could ever be consideredΒ adult-ish. Already we have seen subreddits for smoking cessation, Spotify playlists, Discord channels and even news reports on issues like Gaza, protests and knife crime blocked. Sites such as Wikipedia who are crowd sourced and do not have the resources to age-gate their contents could find themselves being forced to make their content unavailable in the UK, or face fines they have no hope of being able to pay.[4]Β This raises huge conflicts of interests for childrenβs rights β we face the ridiculous prospect of sixteen-year-olds being eligible to vote while being shielded from huge sways of world news. We could have recording artists who will be unable to listen to their own music on streaming services, and teenagers unable to research their EPQ projectsβ¦Β [5]
What is more concerning, from our perspectives, is prohibition of young people being able to access reliable information on topics such as sex, relationships, contraception, eating disorders, suicide, substance misuse, or sexuality and gender dysphoria. This is especially problematic, as many of these topics are likely to be missed by Relationship and Sex Education (RSE) in schools, or even if they are discussed, if young people have questions or concerns, they may be unable to investigate further and may find the answers they seek restricted and age-gated[6]. There is the potential for this policy to become the digital equivalent of the book bans we have seen in American school libraries. This is particularly problematic for those young people under 18 who may be in abusive relationships with their partners or questioning their sexuality or gender identity, but do not live in a supportive home or community. Once again, we are seeing the very provisions supposedly put in place to protect young people putting them directly at risk. Too often young peopleβs rights are framed only in the guise of protection from, rather than access toβ¦ in this case their right to information.[7]
Unfortunately, the focus of debate around age verification has largely been centred around restricting access to βadultβ content under the guise of protecting children. The stated aim of the Online Safety Act is to enhance childrenβs online experiences, not merely restrict them in the digital world. As Sonia Livingstone, one of the leads of the 5Rights Foundation, argues:
βSociety cannot aim to protect children at the cost of restricting their civil rights and freedoms, nor by policies that, however inadvertently, incentivise businesses to age-gate children out of beneficial digital servicesβ[8]
As adults discussing childrenβs rights, we find it much more comfortable to focus on the rights of provision (e.g. health care, housing and family) and protection (from harm, abuse, etc) but seem to forget childrenβs rights to participation. This includes Article 13 β freedom of expression, including the right to seek, receive and impart information and Article 17 β the right to access reliable information. At what point is the cost of protection too great? When considering the best interests of the child, we should do so based on the entirety of childrenβs rights, including childrenβs right to privacy.
Can I see your ID please son�
Age gating content seems like a straightforward idea β it is certainly nothing out of the ordinary and a regular occurrence in the real world, especially for young adults. It is common practice to ask for ID from anyone under 25 years old when buying alcohol, or younger teens when buying tickets for the cinema, or even when buying paracetamol in the supermarket. So why does this not seamlessly translate to online platforms?
There are a number of practical issues β in a bar, or a shop there is a person who will physically look at your ID and then give it back. It is rare for it to be scanned or copied as you go about your everyday business. Obviously, this is not what will happen online β instead, users are asked to upload identification, credit card details, biometrical data or a short video where their face is recorded from various angles to be assessed by AI age estimation tools. This is where the issues arise.Β
As it is left to the platforms to put provisions in place to age gate its visitors, a number of different digital solutions are being implemented, many of which are provided by third party platforms β none of which, it seems, are housed or regulated in the UK. This raises huge data privacy concerns, as many have been known to use insecure servers, and have retained data which they have promised would be deleted immediately. There are huge fears databases housing, particularly data relating to individualβs pornography use, would be identified as honeypots by hackers.
We have seen this happen recently with the hack of the mobile app Tea, which resulted in over 72,000 images β including 13,000 photo IDs of its user captured. Tea is a popular mobile app designed as a ‘dating safety tool’ with the stated aim of protecting women. The women-only app allows users to share accounts of men they have dated, put out warnings of abusive behaviours, and access background checks on men. It is thought the site was specifically targeted after a βhack and leakβ notice was posted on 4Chan, which is largely populated by members of the manosphere. We have already witnessed the first hack since the implementation of the Online Safety Act as a third-party organisation employed by Discord was breached. The messaging app, which is popular with young people in the gamer community, was not itself breached, instead hackers targeted a firm responsible for verifying the ages of its users, with an estimated 70,000 user IDs stolen. This is likely to be only the first of many such incidents.
Another option for platforms to use are AI age estimation software. Users are asked to either turn on their cameras, or to upload a short video where their face can be analysed by AI facial recognition tools which will provide a probably age of the user. However, a recent Australian study, using a data set of 10,139 facial images found significant inconsistencies in classification performance[9]. Notably, the software demonstrates racial bias, age and gender bias, with highest accuracy for the white users and lowest accuracy for black faces. It also displays both age and gender bias, with lower accuracy for young males compared to young females. It has been well publicised in recent weeks that this software can be quite easily tricked as young people have posted on TikTok and Instagram their efforts to fool the tools using characters from computer games, rather than their own faces β to much success.
Indeed, the farce of the situation is the ease of which these new guardrails can be sidestepped. Proton VPN recorded an 1400% hourly increase in signups over the weekend following the implementation, and VPNs became the most downloaded apps on the Appstore. This would suggest this policy is performative rather than practical in any meaningful sense. Whilst, for adults, investing in a VPN subscription from a reputable company might be a sensible, and safety conscious choice to not just watch porn but also protect your devices on public networks, such as when travelling, or working in coffeeshops, young people are not afforded the same luxury. With limited funds, they are more likely to download free VPNS that may be phishing scams or contain malware or instead may be pushed to less regulated corners of the internet where they are even less safe in order to by-pass these safeguards or to protect their privacy.
Indeed, the biggest irony is who is profiting from the implementation of AV. For example, AllpassTrust, is one of many third-party, age gating, tools platforms are utilising. However, with a quick web search we can find AllpassTrust is actually a trading name of Aylo Free Sites Ltd, who are owned by a private equity firm based in Canada; the same equity firm who just happen to own Pornhub (and a number of other adult aggregate video sites). We are now in the ridiculous situation where the companies collecting, storing and profiting from usersβ data, all in an effort to protect children from the very content the same companies are producing!
Developmentally (in)appropriate�
To coincide with the enactment of the Online Safety Act, in August, The Childrenβs Commissionerβs Office released its latest report:Β βSex is Kind of Broken Nowβ: Children and Pornography. In the introduction to the report Rachel de Souza states:Β βThis report should be read as a snapshot of what rock bottom looks like [β¦] As Childrenβs Commissioner, I have consistently said that no child should be exposed to pornography online.βΒ We need to ask ourselves, what do we mean by the phrase βΒ children exposed to pornography? There is a huge difference between a 6-year-old stumbling over an older siblingβs or parentβs age-inappropriate search history β the internet has moved a long way from insistent pop-ups β and a sixteen-year-old searching for porn to masturbate.Β Β The termΒ exposedΒ is doing a lot of heavy lifting here. The problem with this statement, and the report in general, is the language of moral panic. Whilst it may be convenient to lump everyone under the age of 18 together, it confuses the picture.
We would argue, that whilst no one is suggesting we should not protect children from adult content, at some point, children become adolescents and later young adults. Content, however distasteful we may find it, that is completely inappropriate for one age category, at some point becomes developmentally appropriate for another. It makes perfect sense why teenagers may be curious about -nay, actively choose to seek out β pornography. And yet, we do them a disservice to infantilise them under a single umbrella ofΒ exposed children, depriving them of any sense of personal agency and negating their ability to engage critically with the media they consume.
Indeed, one of the essential domains of healthy sexual development identified by McKee et al, (2010)[10]Β is open communication with trusted adults about sex and relationships. Our fear is by employing strict censorship regimes and discourses of harm and risk, we shut down potential conversations rather than opening up space for curiosity and dialogue. In the language of safe uncertainty, by attempting to enforce a position of safe certainty characterised by rigid rules and strict boundaries in an attempt to remove risks, we inescapably drift into positions of unsafe certainty, causing more harm than good. Research evidence suggests notions of harm from exposure to pornographic content, that young people hear from safe adults and the media, bother young people more than the actual material itself[11]. For example, when parents get angry or upset at young people when they stumble across porn, this is more damaging that the sight of whatever they watched. Restricting access inadvertently contributes to a culture of shame about sex that is inherently unhelpful and unhealthy
To put things simply βΒ as Ranumβs law states, you canβt solve social problems with software.Β From the perspective of sex education, the issue of how to ensure young people arenβt harmed by unwanted encounters with pornography is best managed not by restrictions β but by educationβ¦ and creating space for open conversations.
Whilst age verification may promise certainty, in reality it only delivers a false sense of safety: an easily bypassed technical fix that risks suppressing young peopleβs access to support, information and rights, while doing little to address or prevent the behaviours it targets. These restrictions focus on controlling symptoms rather than understanding the needs and drivers that draw young people online. If we want young people to navigate sexual content safely and critically, the solution cannot lie in prohibition alone. It requires education that invites curiosity, dialogue and relational trust, which are the cornerstone of safe uncertainty.
If you would like to know more and better understand how the safe uncertainty framework can be applied to pornography and porn literacy β how education grounded in open conversations rather than silence or shame – can better equip young people to make informed and ethical choices in their digital lives, please read our response to the Pornography review.
[1] It should be noted there is no such offence of βunder-age viewing of pornographyβ in England and Wales. It is a distribution law (similar to alcohol) current law places obligations on platforms, not on the child who accessed the content, while a childβs misrepresentation of their age does not ordinarily constitute fraud.
[2] Blake, P (2019). Age verification for online porn: more harm than good? Porn Studies Taylor Francis Online, Feb 2019
[3] Stardust, Z., Obeid, A., McKee, A., and Angus, D. (2024) Mandatory age varication for pornography access: Why it canβt and wonβt βsave the children.β Big Data & Society, April-June 1-17. Sage.
[4] Crompton-Reed, Lucy and Rebecca MacKinnon. 2023. βProtect the Future of Wikipedia in the UKβ.
Wikimedia Foundation. 29 June. https://wikimediafoundation.org/news/2023/06/29/protect-
the-future-of-wikipedia-in-the-uk/.
[5] βWhat if there was a simple test to make sure every new internet regulation preserved the spaces and parts of the internet that you love the most?β The Wikipedia Test β article by The Wikimedia Foundation 27th July 2025 – https://wikimediafoundation.org/news/2025/06/27/the-wikipedia-test/
[6] Petley, J. (2025): Porn and the OSA: what about the children?, Porn Studies, DOI: 10.1080/23268743.2025.2476961
[7] Third, A., & Collin, P. (2016). Rethinking (children’s and young people’s) citizenship through dialogues on digital practice. In A. McCosker, S. Vivienne, & A. Johns (Eds.), Negotiating Digital Citizenship: Control, Contest and Culture (pp. 41-59). Rowman & Littlefield.
[8] Livingstone, Sonia. 2023. βHow Can we Make the Internet Safe for Children in Practice?β. LSE Comment. https://blogs.lse.ac.uk/politicsandpolicy/how-can-we-make-the-internet-safe-for-children-in-practice/#:~:text=One%20in%20three%20children%20uses,age%2Dassessing%20all%20their%20users%3F.
[9] Stardust, Z., Obeid, A., McKee, A., and Angus, D. (2024) Mandatory age varication for pornography access: Why it canβt and wonβt βsave the children.β Big Data & Society, April-June 1-17. Sage.
[10] McKee A, Albury K, Dunne M, et al. (2010) Healthy sexual development: A multidisciplinary framework for research. International Journal of Sexual Health 22(1): 14β19
[11] SpiΕ‘Γ‘k S (2016) Everywhere they say that itβs harmful but they donβt say how, so Iβm asking hereβ: Young people, pornography and negotiations with the notions of risk and harm. Sex Education 16(2): 130β142.