Growing up in the online world: a national conversation

Supporting children’s wellbeing online is not a one-size-fits-all endeavour. The features, content and safeguards appropriate for younger children will differ from those that teenagers need. As children grow up, the balance between benefit and risk of time spent online becomes more nuanced. As they grow older, teenagers gain more independence in all aspects of their lives and their engagement with technology should reflect that. This chapter considers a range of interventions aimed at supporting children to have safe and enriching online experiences. Some of which are mutually exclusive, some of which could work alongside each other. Across all of them we are seeking views on the appropriate age were they to be applied. Across the interventions discussed in this chapter there is also the question of who they should apply to. There is no straightforward definition of ‘social media’ and many of the things children, parents and carers are concerned about also appear on other services including, but not limited to, gaming, chatbots and messaging services. At the same time, we want regulation to be proportionate and to be effectively targeted to achieve the right balance. We will need to decide which providers are in scope of any new rules, and how to keep this list clear and up to date. We are seeking views on how broadly these interventions might apply, criteria and definitions that should be used if any of these options are taken forward and any exemptions that might be appropriate. The evidence we receive will inform how online services, including but not necessarily limited to social media, might be restricted to build our children’s wellbeing. We will consider the possible impacts of any potential restrictions on vulnerable children, including those with special educational needs and disabilities ( SEND ). We will also consider how any transitional arrangements might work in relation to new restrictions, for example for existing accounts. Restricting social media services by age There is no current minimum age for accessing social media set in law. While the protections of the Online Safety Act apply to all children on user-to-user and search services, the Act only mandates ‘consistent’ enforcement of services’ own minimum ages and highly effective age assurance is not mandated. However, the Online Safety Act already requires platforms that allow pornography, self-harm, suicide or eating disorder content to use highly effective age assurance to prevent any child under 18 from accessing these types of content. Most social media services that set a minimum age do so at 13. This aligns with the age of digital consent – the age at which a child can consent to the processing of their data. The digital age of consent and its role in supporting protections for children online is discussed later in this chapter. Even though many of the most popular social media services set a minimum age of 13, there is extensive evidence that much younger children regularly use them.  Ofcom ’s 2025 Parents and Children: Media Use and Attitudes Report found that 81% of 10–12-year-olds use at least one social media app or site and 86% have their own account on a social media, messaging or video sharing platform. [footnote 7] We consider how enforcement mechanisms could be improved in Chapter 3  but legislation mandating a minimum age of at least 13 to use social media, alongside a requirement for this to be effectively enforced, would be one way to address this. This would equate to a ban for anyone younger than the minimum age. It’s important to note, that this may mean adults being required to age assure to access these services, as a means to ensure effective enforcement of minimum age limits (more on this at Chapter 3 ), which could create trade-offs to user privacy, and create greater friction with adult users’ online experiences. However, depending on how it was defined, introducing a minimum age limit for social media carries a risk of unintended consequences – even if it were set at 13. There are some online services which resemble social media but which are specifically designed for younger children, with features and user-to-user interactions subject to careful moderation. This approach can, when used as part of a balanced set of activities, provide benefits to children. For example, Internet Matters has found that for children between 6-10 years old, children’s use of digital technology can boost language skills, social development and creativity in children. [footnote 8]

Box 2.1. Example : Children’s use of digital technology can significantly boost their language skills in creative and engaging ways. By watching educational videos on platforms like  YouTube Kids , children are exposed to new vocabulary, correct pronunciation, and varied sentence structures. Interactive tools such as  Scratch  encourage them to write dialogue and instructions, strengthening their reading and writing skills.

There are growing concerns among parents, carers and civil society organisations about older children too. Over 180,000 families have signed the Smartphone Free Childhood Parent Pact to delay getting their child a smartphone until at least 14, and social media until 16. [footnote 9] In only 2 years this grassroots organisation has evolved into a national movement, underscoring how deeply parents across the UK are worried about technology’s addictive nature and the need to safeguard children. Some teenagers themselves have said that they worry about the impact of spending too much time online and feeling like they aren’t able to control their own screentime. Additionally, 2025 research from Ofcom shows that a third (33%) of 8–17-year-olds say they have seen something online that they found ‘worrying or nasty’ in the past 12 months. This is unchanged since 2023. [footnote 10] The momentum behind a ban is understandable: it provides a simple, easily communicated approach for parents, children, teachers, those that work with children and young people, and industry. Parents in Australia have been reported to be widely supportive of a ban, agreeing that it would help protect the mental health and wellbeing of children. [footnote 11] Teachers have also expressed strong support, with the Teacher’s Union ( NASUWT ) calling on government to tighten legislation so technology firms would be fined for allowing children under 16 to access their platforms. We understand that parents want action to support their children’s wellbeing now and a ban appears to be a simple way to deliver it. That said, civil society organisations, including the NSPCC , the Molly Rose Foundation and the Internet Watch Foundation have raised concerns that this might not be the right approach and risks doing more harm than good. Their concerns include creating a ‘cliff-edge’ for older teens when they do come online and driving children to use riskier platforms. Additionally, there is concern that introducing a ban for under 16s could make children less likely to seek the help of adults if they do see something worrying online, particularly if they have taken measures to circumvent a minimum age limit. These organisations also highlight the important role that social media can play in providing sources of support and guidance for children, especially for vulnerable children or those who may have more limited access to trusted and engaged adults. It is right that we ask questions about a minimum age, for example of 13 or 16, allowing policy decisions to flow from the evidence we gather. This evidence-based approach to the evaluation of any potential ban is supported by key stakeholders, including the Breck Foundation. [footnote 12] These responses will be considered together with the answers to questions about which sorts of services restrictions should apply to, which is covered in more detail later in this chapter. The debate about introducing a minimum age for social media has raised questions about the potential implications for parents and carers if children circumvent new rules. The government is clear: no parent, carer or child will be fined or punished if these safeguards are circumvented by children.

Consultation questions (4 to 7)

1. Would you support a legal requirement for social media services to have a minimum age of access?

a. Yes

b. No

c. Don’t know/Prefer not to answer

1. To what extent do you agree or disagree with the following statement: “Social media services should have a minimum age of access of at least 16 and should not be accessible to any children under that age”

a. Strongly agree

b. Somewhat agree

c. Neither agree nor disagree

d. Somewhat disagree

e. Strongly disagree

f. Don’t know/ Prefer not to answer

1. Would you support a legal requirement for social media services to have a minimum age of access lower than 16? If so, at what age would you set it?

a. Yes – 13

b. Yes – 14

c. Yes – 15

d. No – not lower than 16

e. Other (please specify)

f. Don’t know/ Prefer not to answer

Please explain the reasoning behind your answers about minimum age requirements.

1. What do you think the impacts would be of having a minimum age requirement higher than 13 for social media services?  For example, impacts on the safety and wellbeing of children, or the impact for parents and carers, as well as other users. You could also comment on the impact on all users’ privacy and data or on business costs, revenue, and innovation.

Age of digital consent Different online platforms may have different reasons for collecting personal data. For example, online retailers may need to collect personal data to remember what items have been placed in an online shopping basket or collect shipping details and credit card numbers to deliver items that have been purchased. Search engines and social media companies may collect behavioural data from their users to personalise and shape their online experiences, including to make recommendations or deliver targeted advertising. The UK Information Commissioner’s Office’s ( ICO ) Children’s Data Lives research shows that children frequently share data through apps and devices, entering personal information, posting updates, and interacting with algorithms. [footnote 13] Under the UK’s General Data Protection Regulation ( UK GDPR ), whenever an online platform is using personal data, it must identify a legal ground for doing so. Depending on the purposes and the nature of the data collection, different legal grounds may be available. For example, some online services may rely on grounds of ‘legitimate interests’ or contractual necessity to deliver a core service that a customer has requested. However, for non-core services or optional extras, such as profiling a web user’s activity for the purposes of targeted advertising, the most appropriate legal ground is consent.  ICO  guidance on this issue recommends that online services give the child (and the parent where appropriate) as much choice over these elements of the service as possible. [footnote 14] Under the UK GDPR , children can only give their own consent for their data to be processed by an ‘Information Society Service’ [footnote 15] – such as a social media, online gaming or online retail platforms – once they are 13. For children younger than the digital age of consent, organisations are required to make reasonable efforts to verify this consent from those with parental responsibility. Examples might include authentication through a registered credit card, or a referral to a parent’s email address asking them to authorise consent. Changing the age of consent would not affect online services that do not rely on consent as their legal basis for processing. Although such services would still need to process children’s data in compliance with the general principles of data protection law, they would not automatically be required to seek parental authority to do so in respect of children under the age of consent. However, for services that currently rely on consent as their basis of processing, it could provide another intervention point that gives parents and carers more control over how platforms are using their personal data. There is an option to increase this age from 13. Research shows that children’s understanding of how their personal data is collected, tracked, profiled and monetised develops gradually across adolescence, and that there is no clear developmental milestone at age 13 that would justify the current threshold. [footnote 16] Adding an extra parental check for children between the ages of 13 and 15 could reduce the risk of personal data being used in ways children in that age group might not understand. Across Europe, several countries have already set higher thresholds: 14 in Austria, Italy and Spain; 15 in France; and 16 in Germany and Ireland. In those countries, social media platforms and other online services that rely on consent to process personal data, will require authorisation from someone with parental responsibility before they can process the data of children below the specified age. However, once consent is granted either by a child who is at the specified age or by a parent or carer in respect of an underage child, the online platform is free to process the child’s data. The platform would still have obligations to process the personal data fairly, lawfully and transparently in accordance with the data protection legislation, but in practice it may not change the way some features and functionalities work, and some services would still seek to monetise the child’s data through personalised algorithms. The options to make changes to these sorts of online environments are set out in the section below on features and functionalities. Additionally, research also shows that raising the age does not necessarily stop teenagers under the age of consent from using social networks. [footnote 17] Currently, many services often default to self-certification mechanisms like tick boxes, that don’t reliably check age or involve parents or carers. [footnote 18] Changing the age of digital consent may only be effective if organisations ensure their verification processes are as robust and meaningful as possible when a child grants their own consent or when parental authority is sought in respect of underage children. This relates to the issues set out in Chapter 3  on age assurance. Parents and carers may have different attitudes about how much personal data of their children online services should be permitted to collect. They may also vary in their levels of digital literacy and ability to engage with parental verification requests. This means that children of parents or carers with less confidence or capacity to engage with digital processes may have reduced access in practice, potentially widening existing digital inequalities. This could lead to disparities in the online experiences of different groups of children. If the UK raised the digital age of consent, it could also limit access to useful services and tools – such as educational technology – that rely on consent to process children’s data, unless some services were exempted. Any change would need to balance stronger protection for children whilst ensuring they can still use beneficial online resources. This consultation aims to understand the level of demand for raising the digital age of consent, the potential impacts of changing this level of consent and how we should ensure any changes are effective and workable.

Consultation questions (8 to 11)

1. At what age do you think the age of digital consent in the UK should be set for information society services?

a. 13

b. 14

c. 15

d. 16

e. Other (please specify)

f. Don’t know / Prefer not to answer

1. What risks or burdens may be associated with raising the minimum age of digital consent? For example, ensuring parental consent, costs to industry, access to services, volume of requests, etc.
2. What should be considered to make raising the digital age of consent effective and workable? For example, suitable approaches to verify users’ ages (including where parental consent is required) or suitable approaches to verify a parent or carer’s identity, age and relationship to the child.
3. To what extent do you agree or disagree with the following statement: “There is a case for changing the digital age of consent for some online services but not others”

a. Strongly agree

b. Somewhat agree

c. Neither agree nor disagree

d. Somewhat disagree

e. Strongly disagree

f. Don’t know/ Prefer not to answer

Please explain the reasoning behind your answer.

Restricting access to services based on features and functionalities Some functionalities and features increase the risk of children being exposed to harmful or inappropriate content. This might be because they are designed for the content to be seen in real-time or on a temporary basis, making it much harder for services to provide protections for children. Examples include livestreaming or disappearing messages. Another way of restricting social media for certain ages would be to identify features and functionalities which are inappropriate or present such risk to children that they outweigh any benefits of use. Data protection law requires organisations that use personal information, including that of children, to do so according to a number of principles – including to use information only in fair, lawful and transparent ways. The ICO ’s Age-Appropriate Design code sets out how online services should provide those services and use children’s data in a way which is age appropriate, including ensuring privacy settings for children are set to high privacy by default. The Online Safety Act already requires online services to prevent children from seeing illegal content, and the most harmful types of content – this includes pornography and content that promotes, encourages or provides instructions for eating disorders, self-harm and suicide. Services are also required to provide children with an age-appropriate experience – this includes ensuring they put in age-appropriate protections for content including bullying, violence and dangerous challenges. As part of this, services must look at whether certain features on their platforms make it more likely that children could come across illegal or harmful content and mitigate the risk if so. For example, companies are now legally required to address algorithms that push harmful content, such as suicide or self-harm content, to children and to take steps to prevent unknown adults from contacting children. Further restrictions on features and functionalities are being proposed by Ofcom , for example the regulator has consulted on protections for children while live streaming. It should of course be noted that there are benefits to some of these features and functionalities for children. For example, hosting a live art session, music performance, or storytelling via a live stream can encourage children to express themselves creatively. Additionally, disappearing messages could cause children to feel more relaxed sharing everyday updates or light conversations, like speaking face-to-face. This can promote natural, informal communication. However, the risks may outweigh these benefits. Government is seeking views on whether any features or functionalities are so high-risk that, even with mitigations, they should be off-limits to children, either at all or below a certain age. This could result, for example, in a set of services with functionalities suitable for children under 16 and a set of services with functionalities which are only accessible to children 16 and over. This would create an easier online world for parents and children to navigate. It would ensure much more transparency about the suitability of different services for different ages. This would deliver similar outcomes to the ‘age classification’ system that currently exists for other sectors, in a way that accounts for the dynamism and scale of online activity. Some online services use end-to-end encryption. The Online Safety Act does not prohibit the use of end-to-end encryption. Services that use end-to-end encryption are already required under the Act to assess the risks their service poses to children and to take proportionate steps to mitigate those risks, including in private communications. We would expect the same principles to apply regarding the measures being consulted on in this consultation. Simply having an end-to-end encrypted service would not be an acceptable reason for a platform not to comply. These features and functionalities might, for example, include:

Livestreaming:   NSPCC ’s 2018 survey of 40,000 children aged 7–16 found that 24% had livestreamed, and 6% of those received requests to remove or change their clothes. [footnote 19] The Internet Watch Foundation ( IWF ) similarly reported engaging with more than 2,000 cases in which children were believed to have been groomed or coerced into livestreaming illegal content – much of which appeared to have been recorded by offenders. [footnote 20] Some services already voluntarily prevent users they know to be under a certain age from accessing these features, for example you must be at least 18 to host a live stream on TikTok.

Ability to send and receive images and videos containing nudity:  The creation, possession and sharing of indecent images of children is illegal under UK law. Over 92% of sextortion cases involved recording or sharing content on social media, [footnote 21] with teenage males aged 14-17 particularly at risk of this harm. [footnote 22] Many platforms already deploy technology to detect the sending and receiving of images containing suspected nudity. In practice, these tools are often used to blur or warn, rather than to block the transmission of illegal content. This detection can occur before a message is sent and does not require access to message content, including in end-to-end encrypted services. The consultation therefore seeks views on whether and how companies should be required to prevent children accessing services which allow the transmission of nude images, rather than relying on voluntary or partial measures.

Location sharing: Many children, parents and carers find these features useful on certain services, allowing children to grow in their independence whilst still enabling parents and carers to check on their whereabouts. However, on other services they can help facilitate illegal harms, including grooming, harassment, stalking, and coercive or controlling behaviour. As information about a child’s whereabouts is personal data, services should be complying with the requirements of the UK’s data protection legislation. Under their Illegal Content Codes of Practice,  Ofcom expects that services that allow users to interact, that assess themselves to be at a high-to-medium risk of grooming, should ensure children are unable to share their location.

Stranger-pairing: Under the Online Safety Act, some platforms have requirements to turn-off functionalities by default where they enable children to speak or connect with strangers. This includes services, that allow users to interact or post publicly (user-to-user services) and have a high risk of grooming on their service.  Ofcom  identifies stranger-pairing features as high risk due to increased exposure to harms such as grooming, harassment, pornography, suicide and self-harm content, hate, and bullying. Evidence particularly highlights risks on gaming platforms, where pairing with unknown users is often normalised as part of gameplay. [footnote 23]

Disappearing messages:  Disappearing messages are a common feature across various social media and messaging services. While intended to promote privacy and informal communication, the temporary nature of these messages can make harmful content harder to evidence or report, and may increase opportunities for coercion, grooming, or peer pressure. Ofcom ’s Illegal Content Codes require services with direct messaging to take into account additional risks posed by disappearing messages to reassure users that that there is no permanent record of the content they are sending. Ofcom ’s Children’s Codes identify disappearing messages as an additional risk factor in children sharing violent content or bullying.

This list is not exhaustive, and we welcome suggestions for other features or functionalities where the risks to children, even with mitigations, outweigh the benefits, and on the appropriate age for any restrictions. We welcome views and evidence through this consultation and national conversation on relevant best-practice industry examples, to help us identify where voluntary company innovation may drive change across the industry.

Consultation questions (12 to 15)

1. Some online services allow their users to engage with the following functionalities. Do you think these functionalities should be age restricted so that children below a certain age cannot engage with them? (Please select all that apply)

a. Live streaming

b. Ability to send nude images or videos

c. Disappearing content

d. Location sharing

e. Connecting or talking to strangers

f. None of the above

g. Other (please specify)

h. Don’t know/ Prefer not to answer

1. Based on your previous answers, please specify your preferred minimum age for each of the functionalities below:

a. Live streaming

b. Ability to send nude images or videos

c. Disappearing content

d. Location sharing

e. Connecting or talking to strangers

f. None of the above

g. Other (please specify)

1. To what extent do you agree or disagree with the following statement: “Restricting children’s access to these features/ functionalities, would provide for a safer online experience for children”. Features/functionalities include live streaming, the ability to send nude images or videos, disappearing content, location sharing and connecting or talking to strangers.

a. Strongly agree

b. Somewhat agree

c. Neither agree nor disagree

d. Somewhat disagree

e. Strongly disagree

f. Don’t know/ Prefer not to answer

1. What do you think the impacts would be if some online services were required to introduce age restrictions on specific features and functionalities? For example, impacts on the safety and wellbeing of children, or the impact for parents and carers, as well as other users. You could also comment on the impact on all users’ privacy and data or on business costs, revenue, and innovation.

‘Addiction’, compulsive design and displacement Research suggests that high levels of technology use can interfere with children’s ability to concentrate and engage effectively in school, [footnote 24] disrupt healthy sleep patterns, [footnote 25] and contribute to broader challenges around mood, behaviour and overall wellbeing. [footnote 26] Phones pose a distraction matter for children outside of school too, whether this applies to when young people are doing homework, over family dinners or spending time with friends. Restrictions on phone use in-school can be a good way of building healthy habits that apply more generally in children’s lives. Ofcom ’s Children’s Passive Online Measurement tool found on average, children aged 8-14 spend nearly 3 hours a day online across smartphone, tablet and computer. [footnote 27] An Ofcom report found that 99% of children aged 13-17 reported that they see at least one benefit to being online, such as helping them with school and homework or maintaining friendships. [footnote 28] The same report found that 33% of 8–17-year-olds agreed that their screentime was too high. [footnote 29] An  Ofcom  report found that 4 in ten (39%) parents agreed that they find it hard to control their child’s screentime, increasing as the child gets to teenage years. [footnote 30]  The National Youth Strategy summary report states that young people need support for both physical and mental health and that they have identified feeling ‘exhausted’ by the digital world. In this section of the consultation, we ask for views on how platform design features may encourage children to stay online for longer. We are also seeking views on the impact of these on children’s wellbeing, especially where they may contribute to the displacement of good mental health, concentration, sleep and self-esteem – all factors that we know are important for a healthy, balanced childhood.

Box 2.2. Evidence around screentime and the impact on children’s mental health

The evidence around screentime and the impact on children’s mental health is mixed, and no causal link has been found. In 2025, a consortium of academics from UK universities, led by Professor Amy Orben, conducted a review of scientific research to understand the impact of smartphones and social media on children and young people. [footnote 31] The review highlighted a lack of high quality causal evidence linking children’s mental health and wellbeing and their use of digital technologies, specifically, social media, smartphones and AI chatbots. However, the report also noted that this doesn’t mean a causal link doesn’t exist and highlighted that there is associative evidence between these issues. This is aligned with the findings of the UK Chief Medical Officers’ 2019 study which also found associations between children’s screen-based activities and negative effects such as increased risk of anxiety or depression. [footnote 32]

Taken in the round, these findings highlight the importance of ensuring that children’s online experiences are balanced and do not come at the expense of their physical, cognitive or emotional development.  Ofcom ’s Children Media Lives study, a longitudinal qualitative research project that follows a group of 21 children aged 8 to 18, highlights some of the negative feelings children in the sample can feel in relation to spending extending periods of time on devices. [footnote 33] This includes difficulties paying attention and feelings of overstimulation. The UK Chief Medical Officers’ suggested that screen use can displace healthy activities, such as sleep. [footnote 34] Adequate sleep can enhance memory, improve school performance and boost moods. For teenagers and younger children, sleep is essential for development. If children are on social media or using screens at night, instead of sleeping, this could be having a negative impact on their development. Technological features and mechanisms can be especially powerful in shaping young users’ behaviour, sustaining long, repetitive patterns of use.  Ofcom research suggests that these include:

Infinite scrolling:  Infinite scrolling means new content keeps loading automatically as you scroll down, so the page never really ends.

Autoplay features: This is a system feature that automatically starts the next piece of content without requiring a user action, once the current content finishes or when it appears on screen.

Affirmation functions:  These are designed to give users positive feedback, validation, or encouragement. These include likes and reactions, comments & replies, shares and reposts, follower counts and subscribers, stories and reactions, streaks, achievements and personalised feedback (e.g. “people are loving your post”).

Alerts and push notifications: An alert is a message that pops up inside an app or website to get your attention right away. A push notification is message sent to your phone or device by an app, even when you’re not using it, to update you or bring you back to the app. [footnote 35]

Additionally,  personalised algorithms  may also play a role in causing children to spend longer online. Content recommendation systems play an important role in social media, along with many other services. They are not inherently ‘addictive’ and can be used to make online experiences more positive and engaging. However, as the experience of Molly Russell so devastatingly demonstrated, algorithms can also be a force for immense harm when they serve children the wrong kind of content, in many cases when they are not proactively seeking it out, or where they drive compulsive use. As the DSIT Select Committee noted in their recent report, ‘Social media, misinformation and harmful algorithms’, thinking about these questions through the lens of the business model can be helpful. [footnote 36] One way of addressing this might be to age restrict services that have personalised algorithms, or to address business models which are dependent on keeping children online for longer, for example by restricting the ability of services to advertise to children or age-restricting features and functionalities that can allow children to make in-service purchases, such as shops or loot-boxes. Parents who are doing their best to support their children to have a balanced relationship with technology may feel overwhelmed by the sophistication of the features and personalisation designed to keep users – including children and young people – engaged. Not all of these functions are inherently harmful but this section of the consultation aims to understand what more can be done to protect teenagers from the functions online platforms use to keep their attention for longer, including:

Restricting access to specific functions for children aged 13 – 15:  This would cover the features or functionalities identified by Ofcom and be kept up to date based on ongoing research. If it is not technically feasible for platforms to remove access to certain functionalities, they may instead choose to bar children under a certain age from using their platform entirely.

Limits on the time children can spend on certain services with these features – “curfews”: Measures here would limit the time spent on social media, set daily screentime limits for individual app usage and/or include nighttime curfews. This would build on measures that some services such as TikTok or Instagram have in place at a voluntary level but make them mandatory and prevent children from ignoring or dismissing screentime limits. Instagram already allows parents to set specific hours during which their child cannot use the platform, as well as allowing parents to make daily time limits mandatory, preventing teens from dismissing these if supervision of their account is enabled.

Nudge techniques to discourage excessive use:  Platforms can voluntarily use nudge techniques to prompt children to take breaks. A recent Danish study suggested promising results – see Box 2.3.

Box 2.3. Evidence box:  A 2025 study by the Danish Competition and Consumer Authority, examined the reduction in time spent online across TikTok, Snapchat and Instagram for 269 teenagers aged 13-17, using ‘nudge’ techniques. [footnote 37] These techniques included forcing a 6 second pause before accessing a platform and asking children to indicate how much time the planned to be on a platform whenever they accessed it. These techniques caused a 31% – 36% drop in usage - equivalent to around an hour a day less for a child who spent 3 hours a day on social media before the study started.

Some voluntary approaches to limiting the time children and young people spend online have been deployed by services on a voluntary basis. This includes on YouTube, where parents’ and carers’ can control how much time their children spend on YouTube Shorts, including setting the timer to zero. Parents can also set a custom bedtime and take-a-break reminders. The consultation will consider whether these sorts of features should be more widespread across industry.

Consultation questions (16 to 20)

1. The following design features are sometimes known as ‘persuasive’, meaning they may encourage children to stay online for longer. From the following list, please select the ones you think are particularly ‘persuasive’ to children: (Please select all that apply).

a. Infinite scrolling

b. Autoplay

c. Affirmation features (e.g. ‘likes’, comments)

d. Alerts and push notifications

e. Content recommendation algorithms (these are algorithms which provide personalised recommendations on a user’s feed)

f. None of the above

g. Don’t know/ Prefer not to answer

h. Other (please specify)

1. Which of these features do you think should be age restricted? (Please select all that apply)

a. Infinite scrolling

b. Autoplay

c. Affirmation features (e.g. ‘likes’, comments)

d. Alerts and push notifications

e. Content recommendation algorithms (these are algorithms which provide personalised recommendations on a user’s feed)

f. None of the above – they should not be age restricted

g. Don’t know / Prefer not to answer

h. Other (please specify)

1. Based on your previous answers, please specify your preferred minimum age for each of the features below:

a. Infinite scrolling

b. Autoplay

c. Affirmation features (e.g. ‘likes’, comments)

d. Alerts and push notifications

e. Content recommendation algorithms (these are algorithms which provide personalised recommendations on a user’s feed)

f. None of the above – they should not be age restricted

g. Other (please specify)

1. Would you support the following restrictions for children’s access to online services? (Please select one)

a. Daily screen time limits for individual apps

b. Restricting overnight access for individual apps

c. Both – Daily screen time limits and overnight access for individual apps

d. I would not support either of them

e. Don’t know/ Prefer not to answer

1. What do you think the impacts would be if online platforms were required to restrict specific features or functionalities, or to introduce time limits? For example, impacts on the safety and wellbeing of children, or the impact for parents and carers, as well as other users. You could also comment on the impact on all users’ privacy and data or on business costs, revenue, and innovation.

What type of services should restrictions apply to Drawing the scope of application for any regulation will always involve trade-offs. For example, organisations such as the NSPCC and Molly Rose Foundation have raised concerns that if minimum age restrictions only apply to a small number of the most popular user-to-user services, this could risk displacing children into other online spaces, including less regulated ones. The scope of application has been a prominent issue of debate around the introduction of restrictions in the Australian system - further explanation is in Box 2.5. The public debate about age limits has focused on social media but there is no single definition of what this includes. The duties in the Online Safety Act apply to search services and services that allow users to post content online or to interact with each other (‘user-to-user’ services). This includes a range of websites, apps and other services including social media services, consumer file cloud storage and sharing sites, video-sharing platforms, online forums, dating services and online instant messaging services. The scope of the Online Safety Act reflects that children can experience harm across a broader range of services than we are looking at in this consultation and relies on risk assessments to identify the appropriate level of protections. The Online Safety Act is already broad and when considering how to build these additional protections for a specific set of risks we will take a more targeted approach. Our aim is to ensure children are protected on services with the features and functions that are most likely to increase their risk of exposure to harmful content, or that have been designed in a way to keep them online for longer. This could include sites and apps that children might not view as being traditional ‘social media’, for example where they might play, create or share games. To target the right services, we will need to set a narrower definition of which ones should be captured. Our starting point is that this should be a sub-set of ‘user-to-user’ services with a number of conditions that determine whether they are in scope. As we consider further restrictions, whether applied to whole services or to specific functions within platforms, we need to meet children where they are. We do not intend for these measures to apply to sites and platforms that are low-use or low-risk for children, for example e-commerce platforms or closed online games that already embed strong safety-by-design protections. This is why we will also be asking about which sorts of sites and platforms that fall under the Online Safety Act should be exempt from any additional new restrictions. However, we are clear that end-to-end encrypted services will not be exempted. Of paramount importance is ensuring that the scope of any new changes is clear to parents, children, industry and the regulator. And given the rapidly evolving nature of technology and services, it will also be key to ensure that any scope is dynamic and future-proof, to allow for the inclusion of new and evolving services. We are consulting on whether the definition should:

align with the Australian definition

include additional conditions based on functionality or purpose

capture services such as some gaming sites and messaging services that might have very similar functionality and risk profiles as those sites traditionally thought of as ‘social media’

include exemptions to avoid capturing low risk and educational services

In addition, we consider issues specific to AI chatbots and Virtual Private Networks ( VPNs ) later in the consultation.

Box 2.4. How the Australia ban works and who is in scope?

Since 10 December 2025, age-restricted social media platforms have been required to take ‘reasonable steps’ to prevent Australians under 16s from having accounts on their service. The restrictions apply to having an account, and Australians under 16 are not prevented from accessing content on these services in a logged-out state, where this is possible.

The Australian regulator, the eSafety Commissioner (eSafety), monitors compliance with the social media minimum age. Age-restricted social media platforms may face penalties of up to $49.5 million AUD if they don’t comply.

Generally, age restrictions apply to social media platforms that meet 4 specific conditions. This is when:

the sole purpose, or a significant purpose, of the service is to enable online social interaction between 2 or more end-users

the service allows end-users to link to, or interact with, other end-users

the service allows end-users to post material on the service, and

the service is not an excluded service under legislative rules.

Platforms that have the sole or primary purpose of enabling messaging or online gaming are among a number of types of services that have been excluded under the legislative rules.

To promote compliance with the social media minimum age, eSafety has recommended that services self-assess whether they meet the conditions. eSafety has published a list of 10 services it has formally assessed to be age-restricted social media platforms, and 10 others considered to not be in scope. Given the number of potential services in scope, eSafety has not published an exhaustive list of all services which are in scope or exempt. Ultimately, any determination of whether a service is or is not an age-restricted social media platform is a matter for the Australian courts.

Social media platforms need to use age assurance methods to comply with Australia’s social media minimum age obligation. eSafety issued regulatory guidance to help platforms understand what ‘reasonable steps’ may constitute for their service. This guidance is principles-based, rather than prescribing specific types of age assurance to be employed. However, the guidance is clear that self-attestation (also known as ‘age-gating’) will not meet the legal threshold, nor will age assurance methods that rely on disproportionate amounts of information being submitted by users.

Consultation questions (21 to 25)

1. What factors are important when determining which apps, sites or services to apply  minimum age of access restrictions  to? For example, user-to-user interaction, the ability to post material, persuasive design features, risky functionalities, the ability to generate non-text mediums such as video or images, the target age group, the size of the service.
2. Are there any types of apps, sites or services that you would want to be captured by  minimum age of access restrictions?
3. What factors are important when determining which apps, sites or services to apply  age-restrictions on specific features and functionalities ? For example, user-to-user interaction, the ability to post material, persuasive design features, risky functionalities, the ability to generate non-text mediums such as video or images, the target age group, the size of the service.
4. Are there any types of apps, sites or services that you want to be captured by age-restrictions to features/ functionalities ?
5. Some services are already exempt from the Online Safety Act. Examples include internal business services, services with limited functionalities and services provided by persons providing education or childcare. Are there additional types of service which you think would be appropriate to exempt from age restrictions? These might include services whose primary purpose is delivery of educational content, services that offer specific child or teen accounts or versions, or services which offer parental controls.

Chatbots and AI Generative AI offers opportunities to improve children’s lives and their education by supporting personalised learning, helping every child develop the knowledge and skills they need and reducing administrative burdens for teachers. AI can also help to drive children’s creativity and imagination, for example helping to co-create stories or inventing characters or worlds. Children might use them as advanced search engines, researching hobbies or skills, particularly on niche subjects.

Box 2.5: The Department for Education recently announced a programme to develop AI tutoring tools to help drive learning and support the most disadvantaged. New Product Safety Standards for education set clear expectations for safe design including: no anthropomorphisation (such as first person or emotion simulating language); no manipulative or persuasive design; no dark patterns; and robust mental health and safeguarding protocols that avoid isolation or secrecy encouraging language. These protections aim to ensure children’s interactions with AI are healthy, transparent and developmentally accurate.

At the same time, evidence on the risks and challenges of children using generative AI is still emerging. It is essential to ensure that AI tools are safe, reliable and appropriate for children. There is growing evidence that UK children are already using AI in their daily life. Nominet polling in 2026 found that 58% of young people (aged 8-17) say that AI makes their lives better, 12% disagree, and almost half (48%) say AI is an important part of their everyday life. [footnote 38]  Barry Laker, head of the Childline service at the NSPCC , said: “AI is becoming a regular part of children’s online experiences. When used safely and responsibly, it can offer opportunities, but it also brings risks, especially if children aren’t sure what’s real, or how AI works”. An event jointly hosted by DSIT and the NSPCC in April will explore further, including with children, the question of the impact of AI on childhood and how it can be harnessed as a positive force. An AI chatbot is an artificial intelligence system with a natural language interface that can replicate human-like conversations with users. They can generate images, videos, text, audio in response to ‘prompts’ from users in real time. They may also remember context and personalise responses to individual users. These services can draw from a variety of datasets including live search results and large language models. AI chatbots can be standalone services or embedded within another service. As chatbots become more embedded in children’s lives, it is vital to understand these risks and ensure their interactions remain healthy and developmentally appropriate. Internet Matters recently found that 64% of 9–17-year-olds use AI chatbots. Of those, 23% have turned to chatbots for advice, and 1 in 8 (12%) say they use them because they have no one else to talk to. [footnote 39] The Online Safety Act provides protection for children from harmful or illegal content on chatbots. Chatbots are in scope of the Act if they either (a) enable user-to-user sharing or (b) search the live internet. Some AI chatbots only draw responses from an underlying model and don’t have sharing functionalities – these are out of scope of the Online Safety Act, except where they are able to generate pornographic content; in this case they would have to comply with the duties to use highly effective age assurance to secure that children are not normally able to access it. The Secretary of State has been clear that she would take further action, including to protect children, where it is necessary. As announced by the Prime Minister on 16 February 2026, the government will introduce new powers so that chatbots not currently in scope of the Online Safety Act can be made subject to duties to protect all users from illegal content. This will ensure the Act keeps up to date with rapidly evolving harms. But some risks stem from interaction with features rather than content. For example, emotional dependence is an emerging harm related to chatbots and one which children may be especially susceptible to. Research on US users suggests it is a significant and worsening risk: long sessions, persistent and anthropomorphic design can lead children to form parasocial attachments with chatbots, sometimes treating them as friends or intimate partners. [footnote 40] Concerns and reports of children coming to serious harm have prompted some services to restrict functionality for children. We want to understand which of the mitigation measures for children proposed in this chapter, for example minimum ages, restricting features and functionalities or time limits may be appropriate to apply to certain chatbots and, if so, how this subset of services is best defined.

Consultation questions (26 to 30)

1. What are the benefits to children of using AI chatbots? For example, this might include as a search function, for educational purposes, for creativity.
2. Which AI chatbot features are most risky for children? (Please select all that apply)

a. The realism of interactions, including realism of content generated

b. The personalisation of interactions

c. How they mimic relationships (friendship)

d. How they mimic relationships (romantic)

e. How they mimic empathy

f. Flattering language

g. Features to encourage more questions/ requests (e.g. asking questions back)

h. The ability to recall interactions across sessions

i. The type of content generated – a) video, b) text, c) audio, d) image

j. Allowing children to have accounts

k. Hallucination or false, misleading responses

l. Ability to engage in and generate mature content (e.g. sexual / romantic roleplay)

m. Other (please specify)

n. None of the above/AI chatbot features are not risky for children

o. Don’t know/ Prefer not to answer

1. Which functionalities of AI chatbots should minimum age restrictions apply to?
2. Should AI chatbots have minimum age restrictions?

a. Yes – minimum age requirements for AI chatbots

b. Yes – restrict access to certain features and functionalities

c. Yes – both minimum age requirements and restricting access to certain features and functionalities

d. No

e. Don’t know/ Prefer not to answer

Please explain the reasoning behind your answer

1. What do you think the impact would be of introducing age restrictions on AI chatbots or certain features and functions? For example, impacts on the safety and wellbeing of children, or the impact for parents and carers, as well as other users. You could also comment on the impact on all users’ privacy and data or on business costs, revenue, and innovation.