Science, Innovation and Technology Committee
Oral evidence: Neuroscience and digital childhoods, HC 57
Wednesday 24 June 2026
Ordered by the House of Commons to be published on 24 June 2026.
Members present: Dame Chi Onwurah (Chair); Tom Collins; Maya Ellis; Dr Allison Gardner; Kit Malthouse; Freddie van Mierlo; Martin Wrigley; Daniel Zeichner.
Questions 39-109
Witnesses
I: Ravi Iyer, Managing Director of the Psychology of Technology Institute, University of Southern California.
II: Damon De Ionno, Managing Director, Revealing Reality; Colette Collins-Walsh, Head of UKAffairs, 5Rights Foundation; and Bernadka Dubicka, Professor of Child and Adolescent Psychiatry, University of York.
Written evidence from witnesses:
Witness: Ravi Iyer.
Q39 Chair: Welcome to today’s meeting of the Science, Innovation and Technology Committee. I thank our witnesses and indeed our Committee members, Clerks and staff who are here in Westminster in the soaring temperatures of today’s heatwave. I want to emphasise, first, that gentlemen and others can take off jackets, and secondly, that if there is a need for a break or if there are any concerns, please raise them; we can always suspend the session if we cannot deal with the temperatures.
This is the second session in our inquiry into neuroscience and digital childhood. We have heard worrying evidence from neuroscientists about the lack of robust causal evidence for the impact of digital devices on children and young people. Today, we are hearing from those who work closely with children about their perspectives on the impact of digital devices, their views on the evidence gap and how policy makers should move ahead even in the absence of clear evidence.
In our first panel, we are very pleased to welcome Ravi Iyer, a researcher who is part of the “Anxious Generation” movement. Ravi comes to us from Boston, where I think it is 4.30 in the morning or something, so while he may not be dealing with the heat, he certainly will be dealing with the time difference. We really appreciate your joining us, Ravi. I hand over to Allison for our first question.
Q40 Dr Gardner: Good morning, Ravi. I want to draw on your expertise as an educational psychologist by asking you to set out your understanding of the relationship between digital devices and young people’s mental health and particularly their brain development.
Ravi Iyer: Let me introduce myself. I am the managing director of the USC Marshall Neely centre. I have also worked with Jonathan Haidt; I am a fellow social psychologist, and I worked with him for almost 20 years. I actually used to work at Meta, so I have some knowledge of many of the design decisions that affect kids. A lot of what I worked on at Meta were design decisions that affect variables that are known to affect the mental health of kids. There are some very specific harms—
Dr Gardner: Brain development as well—I want to look at the actual impacts on the development of the brain as well as mental health more broadly.
Ravi Iyer: Sounds good. I am not a neuroscientist, but one thing that we know about brain development is that it is very sensitive to kids’ behaviours. When kids’ brains are developing during puberty, when they are learning how to navigate the social world, the things they do frequently are going to affect the development of their brains. One study in the United States, showed that kids were on social media for five hours a day. In addition, 45% of kids say that they use these products too much, and 45% also say that it affects their sleep.
We know the effects of using these products so much and we know the effects of lack of sleep on both mental health and brain development. Although I cannot tell you the specific effect on a specific region of the brain, we know that lack of sleep and this sort of overuse have negative effects on both mental health and brain development. Those are just two of the harms. I would also like to talk about the content kids see and the contact they get on these services.
Q41 Dr Gardner: How does the impact vary between different uses and different types of children? Is there variation, or is there a standard trend across all types of use and children?
Ravi Iyer: Certain experiences are unfortunately common across all kids. About 20% of kids in one study said that they see content that makes them feel worse about themselves. There are studies within Meta showing that if you design a platform one way, with like counts and things like that, you get more negative social comparison. We know the effect of negative social comparison—seeing people whose lives you want and whose bodies you admire—on kids’ mental health.
There are kids who are more susceptible to that. Some harms are more pronounced in young women than young men. You tend to see negative social comparison more among young women. Bullying and things like that tend to be reported more by young men. There is a commonality across groups that these are unfortunate experiences that many have.
Q42 Dr Gardner: That seems very focused on social media. I am looking at a wider range of digital use and digital devices. What are the main pathways through which digital technologies, for example educational technology and other uses of technology, affect children? We will come to social media later. I am also interested in the evidence for saying that this impacts mental health and brain development directly. The book “The Anxious Generation” conflates parenting techniques and the use of digital devices. I have asked you two questions in one: how the different types of digital technologies affect kids and what the direct causal evidence is that that is what is harming children and not, for example, parenting.
Ravi Iyer: On the common pathway, beyond social media, I would say that the original sin of Silicon Valley is assuming that the more you use a product, the more value you get. It is not just social media; many products are optimised to be used as much as possible. That leads to overuse. A significant number of kids say that they are use these products too much. They are reporting the negative impact on themselves but are still unable to stop using it.
That does not just go for social media. When I was growing up, I used to buy a video game for $50 and I could play it as long as I wanted. Nowadays, video games try to immerse you in them as much as they can and get you to use them as much as possible so that you buy more things within the game. A user is no longer capturing the excess value. The company is trying to get you to use it more and more.
It is the same thing with gambling. I do not know how much of a problem this is in the UK, but it is a problem in the United States. Brain development does not finish until the mid-20s. Young men are still developing their self-control until then. Gambling companies are sending them the right offer at the right time with AI technologies to get them to gamble more than they want to.
There is a common thread of technology and AI being used to get kids to do things more than they want to. That runs across all the technologies. You see overly sycophantic AI chatbots trying to get kids to use them more than they would otherwise. That is the common thread across all these technologies.
On your second question around the evidence, it is hard to do a random assignment where you assign a society to have or not have a technology. You get much more direct evidence on harms, such as lack of sleep, and no expert would say that lack of sleep has no effect on mental health. There is plenty of evidence that people can talk about; there are experimental studies and longitudinal studies about mental health and technologies, but the best evidence is the direct evidence, which almost nobody disagrees with, about the effect of these products on things such as lack of sleep, and the uncontroversial assertion that lack of sleep has a direct effect on mental health.
Q43 Dr Gardner: Thank you. Lack of sleep is the issue; the cause of the lack of sleep can vary. I am trying to make sure that the pathway of causality is established.
Q44 Tom Collins: Hi Ravi. We heard in our earlier sessions that there is limited causal evidence of the impact of digital devices on young people. Is that your position, too?
Ravi Iyer: That is not my position. I am aware that there are researchers who fall on both sides on the overall impact, and that is where I often focus. I was a project manager at Meta. It is hard to run an experiment on mental health, because the effects are often not seen for a long time. It is much easier to run an experiment on, for example, whether a product intervention or a policy change affects the amount of sleep that kids get. Are they getting notifications late at night? Are they using the product between midnight and 5 am?
Those effects are uncontroversial. They are widely reported both within companies and among teens themselves. Those are the things that one can anchor on and one can measure. One can effect good policy, and I am excited that the UK is considering age limiting accounts on social media platforms. A lot of times, that lack of sleep is facilitated by, for example, notifications sent overnight, which is something that you need an account to have. So that is definitely not my opinion.
Some of that controversy is about the average effect, but I do not think you have to worry too much about the average effect. Let us imagine it is affecting 20% of kids. In the United States there are 40 million teenagers. 20% of 40 million kids is 8 million kids, so a lot of kids are being affected, even if it is not 55%. I am definitely of the opinion that there is clear causal evidence that it is affecting a large number of kids.
Q45 Tom Collins: You are talking about the large scale—very large numbers and statistically high confidence in those trends. I am interested in digging down a bit deeper, because you can assert that social-level devices equals these outcomes, but as we go down and look at the actual characteristics of the device itself, you have talked a bit about how manufacturers are trying to maximise use and you have talked a bit about push notifications—all these reinforcements loops that are trying to be established. I am interested, if you can help us, in trying to break down whether we can draw a clear picture of causality between those characteristics, how we can point to hazards on devices and how that links all the way up to these outcomes that you have been talking about.
Ravi Iyer: If you look at the court cases recently adjudicated in the United States focused on addiction, if you look at the age-appropriate design codes, which are moving from a broad duty of care to illuminating specific features, you will see that they rhyme.
The features that people are talking about in all these cases are similar, and they all relate to increased use of products. They include things such as engagement-based algorithms—can I find the exact video that gets you to stay on a little longer than you want to? They include frictionless interfaces—things such as infinite scroll and autoplay, which do not give you the chance to reconsider the use of these products. That is not just about kids; adults can identify with this, too.
You can also think about things such as notifications and excessive notifications, as well as gamification—a Snapchat streak or how many days in a row. You see kids saying they are not talking to other kids because they want to, but because they want to keep up their streak. This kind of gamification and social comparison, as well as the frictionless interfaces and the algorithmic amplification of things based on AI predictions of what will get people to use a platform longer, is a common feature that is causally linked to people using these platforms more. That is why companies do it. We can definitely isolate those product choices and affect them through good policy.
Q46 Tom Collins: Fantastic. You mentioned a few of these specific design affordances, like infinite scroll and gamification characteristics, such as streaks, as well as learned user characterisation and adaptation of the system. For which of those design affordances is the evidence chain the strongest? How might we build that evidence chain, given that it is very hard to isolate these factors and to experiment?
Ravi Iyer: Platforms will point out that individual features may exist elsewhere. In the US, we have systems that alert you to earthquakes or kids being kidnapped—fortunately, those are rare occurrences. Those may notify you. It is not about that feature in isolation, and it is very different to apply that feature on top of engagement-based algorithms and an infinite amount of user-generated content. It is not just the individual feature. It is the combination of features that are all designed by very smart product managers to get kids to use these products as much as possible. They are actually designed to get adults too—anyone—to use these things as much as possible; it is just that kids are still developing their self-control and are more vulnerable to these systems.
The evidence base is fairly clear. You can look at some of the court decisions in the United States that have adjudicated that evidence. You want to address the combination of these features. I know that the UK is considering an age limit policy. A great way to start is by presuming that the platforms that have this combination of features are harmful until they remove these features, rather than trying to pick out these features one by one.
Q47 Tom Collins: Thank you. You have talked about trying to reduce the hazards. I am interested in how you might demonstrate that these products are safe. You have mentioned age, but we can think about other vulnerabilities—neurological, psychological or social. How might we move from a situation where we are just trying to turn off or restrict things we think are hazardous to a point where we can demonstrate the safety of one of these devices? What is the evidence gap when it comes to defining safety?
Ravi Iyer: We know a lot more than we used to. Some documents about internal product experiments have come out in the court cases. There is a fairly developed evidence base about a set of features that we know are manipulative and addictive, and 70% of kids in the United States say that they feel manipulated by these platforms into using them more than they want to. Probably most of us would say that we use our phones more than we want to. There is a set of features that we have a good evidence base for.
Product managers are going to come up with new features. Just demonstrating that you do not have these features is a clear way that you could demonstrate safety. Just like a food manufacture would demonstrate that something does not have things that are known to cause cancer or other problems, a product designer should also demonstrate that a product does not have features that are known to be designed to cause harm to kids.
Product managers are always going to design new features that get people to use these products more, and I have cited some stats about kids’ own reports of their experiences. Ofcom also does great work in understanding the experiences of kids. We need to try to understand, at a platform-specific level, what kids are saying they are using too much. We did one study recently, which dovetailed well with Pew, that showed that if kids have to pick which platforms affect sleep, TikTok is the platform that is cited the most. We need to understand the specific affordances of a platform and which platform is really affecting kids, and then have a policy that, for example, age limits that platform, so you can see that kids then sleep more. Measuring that would be a way of getting a causal chain that gives you confidence that the policy is going in the right direction.
Tom Collins: Brilliant. Thank you very much.
Q48 Maya Ellis: Hi, Ravi. You have already heard some challenge about the perceived lack of causal evidence in this space. Do you see the recommendations that you have made so far as a precautionary approach in advance of more causal evidence, or do you think that we have enough evidence? Are you comfortable, based on the work that you have done, that we have enough evidence to say that we should make these big decisions?
Ravi Iyer: You will notice that I often frame things in terms of the immediate harms. I do not think there is any controversy about the effect of these platforms on lack of sleep and negative social comparison. In one internal Meta study, 20% of kids reported seeing content that made them feel worse about themselves, so I do not think it is controversial to say that negative social comparison affects millions of kids. I also do not think that the contact effects—contact by strangers, which has led to tragic suicides in some cases—are controversial at all.
When you say there is a lack of evidence, it is really about what dependent variable you are using. I think all these variables point to clear evidence that it affects mental health, but I know experts fall on both sides of that question. However, I do not think experts fall on both sides on its effect on sleep. I do not think any experts dispute that many kids are being contacted inappropriately on these platforms, or that they get social comparison—all the experts would say that that affects maybe not most, but a large number of kids. Yes, we have the causal evidence to address platforms that are causing these industrial-scale experiences for our children.
Q49 Maya Ellis: The reason this is important for us, as Government officials or MPs, is that we have to decide where to invest money in policy or to influence where money is being spent to make a change to this. It is critical that we understand the absolute causes so that we can fix them. In your work, what assessment have you made of the push and pull factors? We are talking about the pull factor of addiction to social media, but in the UK, we have also had a reduction of about 70% in funding for youth services over the past 10 to 15 years, which I would argue is a significant factor pushing people to do things on screens and social media.
Presumably, by extension, there are scientific arguments—or hypotheses, at least—that sleep is affected if children do not have the opportunity to play sports during the day. They have not had enough exercise, so they are not getting enough sleep; we know that a lack of exercise can affect sleep, too. What assessment have you made of the push and pull factors? Can you tell us what is pushing people to social media, as opposed to just the effect of social media or screens?
Ravi Iyer: In my view, there is no tension between addressing both of those things. You do not have to decide to address either the lack of exercise or social media—you can address both. In the case of social media and technology—I was, for a time, a product manager for Meta—platforms run A/B tests, which are scientific experiments about designing a platform in one way or another. It is hard to design a society by running an A/B test on whether the whole of that society does or does not have social media in order to see what its effects are, but it is very easy to design a study about one feature of a platform.
We have had evidence from whistleblowers in some of these cases about the optimisation of platforms for engagement. Specific features of these platforms, which I have mentioned, are the pull factors: they do cause kids to lose sleep and to use them more than they want to, and do cause unwanted experiences, such as seeing sexual content that they do not want to see. There are lots of pull factors on technology platforms that are known and scientifically shown. If experiments are the gold standard of causality, we have experiments showing that product design choices do pull kids into these negative experiences. That does not preclude us from addressing exercise and other things—we should totally address those—but we should also address the pull factors where these platforms are affecting kids causally in this negative way.
Q50 Maya Ellis: I get all that, but arguably, you could do an A/B test in similar demographic communities. In one you severely restrict social media use, and in the other you could drastically invest in local youth services, sports and activities. You could see which one has a bigger effect on reducing social media time. At the moment, it feels like we are talking a lot about the restriction side, but we are not talking about investment and the impact it could have. Do you agree that that would be a decent A/B test—taking similar demographics to see which is the more effective use of public money, to make the shift that we want to see in young people?
Ravi Iyer: You would learn something from that. No experiment is perfect, but you would definitely learn something from it. Kids who live in one community are still on platforms because global celebrities happen to be on those platforms, and those celebrities are global because of network effects that platforms have encouraged—kids have to sign up to those accounts in order to access those celebrities. There are all sorts of network effects that are not isolated to one community, which would make it difficult to do that test.
If the question is, “What is a better use of public money?”, as far as I know, the age restrictions on social media platform accounts are in many ways costless. There might be some cost of enforcement but there are also lots of benefits, such as the mental health benefits that a community reaps. Many court cases in the United States have focused on recovering the costs from platforms. I do not think there is a tension, in terms of money spent, between restricting social media platforms and spending money on other services. Both should be happening.
Q51 Chair: To build on that point, on Friday I visited Hawthorn primary school in my constituency, which was getting a new garden. The first thing that the 10 and 11-year-olds wanted to talk about was not the beautiful new garden that they could play in, but why I was banning social media. Can you answer them? Why should the Government be banning social media? Do you agree with it?
Ravi Iyer: First of all, I want to clarify that I call it an age limit on purpose, because it limits accounts on these platforms. You can still watch what you want on YouTube, and TikTok is surprisingly usable without an account. It is not restricting access to content; it is restricting the business relationship. That business relationship is actually already restricted for users under 13. It is not surprising, but it is worth acknowledging that having 10 and 11-year-olds asking about this illustrates the problem that you get younger and younger kids on these platforms because there is no one enforcing any restrictions.
I am sure there is some age at which it is inappropriate. I have outlined some of the manipulative features. I have also outlined how many older teens feel that it is manipulative. How much more is that the case for someone who is 10 or 11? The average age of pornography exposure in the United States is 12—that is an older study; maybe it is even younger now. A lot of those kids say that they are not looking for pornography, but they see it unwanted. Younger kids especially are not looking for that kind of information, but they are getting it pushed to them by algorithms that are trying to get them to use a platform more. Sure, they will pay attention to it, but that does not mean that they want it. I think there is definitely an age. Can a 10 or 11-year-old understand that when a stranger DMs them, it is not real? The fact that 10 and 11-year-olds are asking about this is illustrative of the problem.
I would ask them why they want social media. A lot of times, kids want these things because other kids have them. That is why it is a Government problem and a collective action problem. That is why it is a problem among parents, who do not want to isolate their kid. They do not want to be the one parent who does not let an 11-year-old on a platform when all the other 11-year-olds are on it, even when all the parents know it is totally inappropriate for 11-year-olds to be on that platform. With 11-year-olds, I would ask them why they want to be on a platform. A lot of them will say, “I want to be on it because all my friends are on it,” and that is the problem we are trying to solve.
Q52 Chair: I did ask them why they wanted to be on it, and they wanted to be on it for social reasons—to communicate with their friends and to be part of a group—so I think there is something in what you say. Can you clarify that you are for a social media ban—or an age restriction, as you call it—for under-16s?
Ravi Iyer: Yes, I am. I have talked a lot about design changes, but, functionally, I would tie it to design features. I point you to a law that is enacted in Indonesia and proposed in Canada, where the presumption is that the platforms are harmful. We have the data, and that is what we have seen and what the kids are saying. Then, the platforms have to prove themselves safe by showing that they do not have certain design features.
Q53 Chair: My understanding from your earlier evidence was that you were considering the ban as something that could be temporary until the social media companies show that their product is safe for children.
Ravi Iyer: I have a lot of reservations about these large platforms ever reforming themselves in a way that would make them safe, but yes, I would tie it to the design features. If platforms did not have harmful design features, then yes, I think we could deem them safe.
Q54 Chair: Why do we not simply legislate now to get rid of those harmful design features, so that the product is safe for everyone?
Ravi Iyer: Because historically, that has just not worked. They are functionally equivalent. It is just a question of where you put the presumption of proof. Do you say, “You guys are unsafe, so you have to prove you’re safe before you can serve kids”, or do you say, “You guys should reform these design features”? Historically, platforms will argue about those specific design features. Introducing an age limit tied to harmful design features just shifts the burden of proof from policy makers to the companies.
Q55 Chair: Basically, you do not believe that the state has the capacity to effectively legislate for what needs to change in social media companies. Is that what you are saying?
Ravi Iyer: I would say that the harms are so clear right now that presuming that they are unsafe and then forcing the burden of proof to be on the companies is—eventually, the state could achieve the goals it wants to in other ways, but historically, looking at Australia’s law, you will move much faster and see much more harm reduction if you put the burden of proof on the companies.
Q56 Chair: I understand that, but my concern is that while you are banning under-16s, the harms that social media companies inflict on those over 16 remain. Social media companies might simply improve for under-16s, but the general harms in the population will remain.
Ravi Iyer: That is not my experience. I will make two points. I still live in silicon valley, in the bay area. I still keep in touch with people across companies, and they say they are able to get through design features that affect not just under-16s but all users more aggressively because of the pressure from these age-limit policies.
The second thing I will say is that, in jurisdictions where protections have been passed for under-16 users, such as in the United States, there are people who used to work on age limits or design changes for youth, and they are now working on scams for the elderly, because it is the same principle. Once you prove that these design features are harmful for kids, it is natural to also show that these design features are harmful for others. I actually do not think there is any tension between age limiting these platforms for under-16s and putting pressure on companies in terms of design. I think you can do both, and in practice, they have been complementary in every other jurisdiction that I have experience in.
Q57 Chair: I am very concerned about the lack of causal evidence, but I take your point that, in some ways, it is the social media companies that have been experimenting on our children. They have the data to do so, because they have access to their own data on who is using their services and on what attracts more use and attention. Is it not really important to get better access to social media companies’ internal usage data and require greater transparency from them, as this Committee recommended a year ago?
Ravi Iyer: Yes, I think that is 100% important. I think companies are doing less of that internally, unfortunately, as a result of the pressure that is being put on globally—that is sort of a natural response. There needs to be external research as well, so maybe transparency can include the ability of external folks to conduct their own research. But I would also say that there is not a tension.
Q58 Chair: Which features would we need to give external researchers access to in social media companies such as Facebook? Can you recommend which key features we should require transparency on?
Ravi Iyer: One short answer would be the product experimentation data they use to make product decisions—any A/B tests that they run—and understand those effects.
There is a second—and actually better—answer: in Utah, they passed a law about interoperability. One way to have experimentation is to have more platforms out there, so that if your favourite celebrity is on Instagram, it is not just on Instagram that you can get that information. Maybe you can build a platform that aggregates information across services. There you will see a lot more experimentation from companies that are a lot more open, and then you are not trying to force them to get that data.
So, one answer is A/B experimentation data, but I would say that another is interoperability. I would look at the Utah Digital Choice Act. Enabling experimentation in the market is actually where you will scalably learn, rather than relying on the right policy to get these large platforms to give you their information.
Q59 Chair: That is very helpful. Is there also the opportunity to see the input and output data of the social media companies? They do not give open feeds or apps.
Ravi Iyer: If you had interoperability, you would get some of that. There used to be a tool called CrowdTangle, which was very influential, externally and internally—you could enable and support that sort of functionality. Knowing the top 100 things that are circulating on a platform can tell you a lot about the incentives of those platforms and how their algorithm is built. If you are not proud of the top 100 things that are on your platform—oftentimes, Meta was not embarrassed by some of those things—you can think of those as bugs that you need to fix in the system.
Q60 Martin Wrigley: Following on from the Chair’s questions, you spoke earlier, Ravi, about platforms getting kids to do more than they want to do. Normally, we call that grooming; it is a bit like how Jesuits were training children under seven to become lifelong devotees. Age limits restrict use for children, and I hear you say that they are perhaps changing how the companies are reacting, but should we really be regulating for a change in behaviour of the victims—we talk about these products almost as if they are an addictive drug—rather than regulating the product producers and putting the responsibility where it really is?
Ravi Iyer: As I understand it, no age-restriction law puts any penalty on any user; they are all penalties on the companies. Companies, again, have their own policies about under-13 users; they are just not enforcing them. That is evidenced by the 10 and 11-year-olds who are asking for access to the platforms. I 100% agree that the responsibility belongs with the platforms. As I understand it, with age-limit laws around the world, there is no fine for any kid or any parent. The fines are for the platforms to enforce their own policies about under-13 users and to raise that age limit.
Instead of regulating the behaviour, I would think of it as regulating the contractual relationship. Most of what kids do on these platforms is consume content passively. It is really about, “I sign up for an account in order to get access to my favourite celebrity, and I am giving you access to all my data. I am consenting to all these design features that I maybe didn’t want. I want to talk to my friends, but now I am getting messages from strangers.” We do not let kids sign up for credit cards; we do not let them enter many contractual relationships that have long-term implications. I would think of it in those terms, as opposed to thinking of it as regulating the behaviour of kids.
Q61 Martin Wrigley: If it is a contract, are you suggesting that we look at the age of contracting, which is typically 18 rather than 16?
Ravi Iyer: I would be thrilled if you looked at 18. We are trying to set a global norm. Many other jurisdictions have aligned on 16, which is an age when a lot of kids have at least exited puberty, though there is still a lot of brain development to go. Looking at the age of contracting and thinking in those terms is perfectly appropriate. In every country, there is an age where you let kids do things that they know are harmful to themselves. They can sign up for the army and they can gamble; they can do a bunch of things that we know require some trade-offs between the long term and short term. We say, “You are old enough to make that decision.” I do not think we have thought of that for social media.
Q62 Martin Wrigley: But in addition to that, we regulate the harmful effects of addictive drugs.
Ravi Iyer: Yes. I am not an expert in addiction, but two signs of addiction are that I use something more than I want to and that I am aware of the negative impacts on my life but I still use it. We see that at fairly high levels for kids on social media; 40%-plus of kids say, “I use it too much and it’s affecting my sleep.”
Q63 Kit Malthouse: One of the arguments that is put against a ban is that children will just get round it with VPNs and all that kind of stuff. These companies are incredibly smart and clever. They are at the frontier of technological development. It is perfectly possible for them to identify a user who is below a certain age, even if they are coming in via a VPN, and put in their way the barriers required. Is that your view as well?
Ravi Iyer: I am actually sympathetic to the fact that it is not perfectly possible to identify every kid, although I think they could do a better job. You can think of making that kind of prediction as having two kinds of mistakes: false negatives and false positives. You incorrectly let a younger kid in or you limit someone who is older and should have access. In any prediction task, there is a trade-off between those two things. I think Australia made the right trade-off. They were a little more lenient. I would suggest going slow as a policy. You don’t have to have perfect compliance on day one. It took us a decade to get into this situation; we shouldn’t try to solve it in two months by having the most draconian policy possible, where we are starting to see older kids and even adults tagged as having under-age accounts, because that will create headaches.
A lot of kids don’t really want to use these platforms in the first place—they might sign up because their school is posting information on these platforms—so you might ease into the policy. The good thing is that age verification is getting better. Apple and Google are constantly coming up with better solutions to make the age verification issue better. So I am sympathetic. They will not be able to identify every under-age user on day one. We want to make sure they 100% comply, because there will be mistakes made. I would suggest going slow and not judging it by how many kids are off in month five, but thinking of it as a long-term process. If we can get what Australia did, which is a great success, with 30% of kids off in the first four months, that is great. That is millions of kids who are already being helped.
Q64 Kit Malthouse: In practical terms, what would you say are the top three most pernicious features that you would require legislation to design out immediately?
Ravi Iyer: Engagement-based algorithms is probably the No. 1 thing. You see it not just on social media, but across services. I would say alignment around the frictionless interfaces, the infinite scroll, the autoplay. Then I would say the gamification features and the things that allow for comparison: things such as like counts and Snapchat streaks. If I can add a fourth, maybe ephemeral content—this idea that if you do not see something in the next six hours, it will disappear and you cannot see it ever. That is also clearly manipulative. Those would be my top four.
Q65 Chair: Thank you, Ravi. To push back on the evidence issue, there has been some criticism of “The Anxious Generation” for the academic rigour of its research and the fact that it speaks to the lack of causal evidence. I have here a quote from a professor of human behaviour and technology at the University of Oxford: “Extraordinary claims require extraordinary evidence. Right now, I’d argue he doesn’t have that.” What is your response to that?
Ravi Iyer: Actually, I think taking up this much of our kids’ attention should require extraordinary evidence of the safety of this product. The actual experiment being done is a giant experiment on our kids’ attention, and I think we should have more of a precautionary principle. In the US, if you try to change school lunches, it requires a ton of evidence, but somehow we have allowed our kids to get on social media for hours and hours each day and there is no evidence about its safety. So I would flip that on its head a bit.
I am aware of those debates. I am not the right person to debate that part of “The Anxious Generation”, but there is a reason why I focus on the intermediate effects—the effects on sleep and of negative social comparison. I do not think those are controversial. I would be curious if you asked the author of that quote what percentage of kids they think are losing sleep, and what percentage of kids they think are engaging in upward social comparison in ways that are harmful. I am pretty sure they would agree that a large number of kids are being harmed and therefore I do not see where the argument is that we do not have a problem.
Q66 Chair: You talked about spending hours and hours on screens. Obviously, banning social media for under-16s does not necessarily reduce screen time. One study found that reducing the time children spent on screens to less than three hours a week made a meaningful difference to emotional difficulties, but that is actually at the bottom quartile of what eight to nine-year-olds spend online. They spend about 300 minutes a week online in this country, so how do we address the screen time question?
Ravi Iyer: It is a part of the solution. You have to start somewhere. You can’t let the perfect be the enemy of the good. We know that, if you ask kids, social media, and in some ways TikTok in particular, is affecting their sleep. You can actually ask them, “Which product is affecting your sleep the most?” and address the product that is affecting their management of time, crowding things out and taking up their time the most. Just because we cannot get all the way to three hours does not mean we should not do that. Just because you do not have one policy bullet to get all the way to under three hours does not mean you cannot make progress.
Q67 Chair: I agree with that, and I am not saying that three hours should be the aim, but I do think—this probably goes to Maya’s point—that removing a certain number of social media platforms will not necessarily mean that children’s screen time will go down. It may mean that they will spend more time on other sites such as YouTube, where you are not required to have an account. That is a displacement issue. The kids at Hawthorn primary school in Newcastle did say, “Why are you regulating our behaviour? Why are you regulating what kids can experience? Why are you not taking action against the bad guys?”
Ravi Iyer: Just to finish my thought earlier, I do think that we have seen meaningful drops in Australia in terms of usage. It has not been 100%, but it has certainly been meaningful.
In terms of taking action against kids versus against platforms, there is no penalty for kids; the penalty is on the platforms, so I do not think it is appropriate to think of it like that. I think it is a bit of a messaging thing. Some call it a ban, but I call it an age limit, and I also try to focus on account creation. Kids cannot sign up for a credit card or for the military. A little bit of this is just how it is communicated, not the reality of the situation. They can still go on YouTube, for example.
Chair: Thank you very much for spending this time with us, given that it is very early in the morning for you in Boston. It has been very illuminating. We really appreciate your research and the evidence that you have given to our Committee today.
Examination of witnesses
Witnesses: Damon De Ionno, Colette Collins-Walsh and Bernadka Dubicka.
Chair: Welcome back to today’s session of the Science, Innovation and Technology Committee, in which we are looking into neuroscience and digital childhood and looking particularly to those who work closely with children for their perspective on digital devices and the evidence gap that we heard of in the first part of our session. Welcome to our witnesses. I am going to ask Freddie to kick us off.
Q68 Freddie van Mierlo: Thank you, Chair. I will ask the starter for 10—and could everyone introduce themselves when answering? What patterns are you seeing in usage of digital devices and social media among children; what, in your various types of expertise, is the impact; and how do you think that is changing?
Damon De Ionno: Hello. I am Damon De Ionno, the managing director and one of the owners of Revealing Reality. We are a social research agency. We cover a lot of different subjects, but for the last 11 years we have been delivering Ofcom’s qualitative Children’s Media Lives research. We have also been doing the ICO’s Children’s Data Lives research for the last three years.
Over that time, we have probably done at least 50 studies into how young people use social media in different ways. That includes more serious harms, like the use of pornography or image sharing, and self-harm and eating disorders, but we also spend a lot of time—I would say we are probably the leading experts—on what the average child does online in a day. In relation to the way they use social media apps, we look at the apps, how they work, the content and children’s specific usage. We track, in some cases, their usage second by second, so that we know exactly what they are doing and at what time they are doing it—yes, I suppose we are experts in that. We spend time with children and their families. We film them. We collect recordings of the content and streams that they see. That is our expertise.
In terms of the trends, when we started doing Children’s Media Lives, children obviously watched a lot of TV and some of them still read books. There have been quite a few shifts in social media over that time. I want to stress that social media has to be divided into messaging and communication, and content consumption. Obviously, many apps have both, but if we are talking about content consumption, which I think is what is behind the question, then yes, there has been a massive move towards short-form content. TikTok is probably the dominant provider of that at the moment.
In the last five years, that content has changed as well. It has become much more commercial and professionally produced. There is a lot less of children making their own funny content and sharing it with friends. I would say that messaging and communication as an activity, and content consumption, have become much more divided in that time. A lot of the content consumption is not that social. It has cultural and social currency for children in the same way as some of us may have felt excluded if we did not have a TV and did not know what programme other children were talking about at school, but in reality every child’s feed is completely different—unique—so it is quite an isolating technology in that sense.
Q69 Freddie van Mierlo: You mentioned that you are looking into how children use social media, but how is that different from how adults use it? I know how I use it, but I have no idea how an 11-year-old uses social media.
Damon De Ionno: We also study how adults use social media, and I can say that the way that you use social media will not be the same as the way that most adults use social media. That is actually a big problem in the policy space, because you will use it as a tool to give you the information you need to do your job. It will be something that assists you and potentially even gives you superpowers when compared with what would have been the case 20 or 30 years ago—the technology is powerful.
Children could, in theory, use social media to do that as well. Such content exists, and I see TikTok at the moment advertising that it helps children learn for exams. That is totally possible, if you went out there to curate a feed that was educational, helpful and positive, but I compare that to letting children run wild in the supermarket overnight and wondering why they all end up in the sweets aisle making themselves sick. It is possible, but it is not the reality for most children.
I would also say that there are demographic differences, and that again is an issue because most policy makers’ children will be having a much more balanced life, and the technology will in some ways facilitate those other activities. I imagine it still takes up more time than it should and that the return on investment is not great—there are other risks as well—but when we look at children who are less privileged, more mainstream or underprivileged, the dominance of social media and the harms are much more pronounced. Obviously, we should be making policy for those people, as well as for people who sit in this room.
Q70 Freddie van Mierlo: I will put the same first question to you, Colette. Could you introduce yourself, say what your organisation does, and then answer what impact you see on children of devices and social media usage?
Colette Collins-Walsh: Absolutely. Good morning, I am Colette. I am the head of UK affairs at the 5Rights Foundation. We advocate for children’s rights online, for all children from birth until 18. We work with regulators, policy makers, industry and children themselves on the technical solutions to how to build this with children’s best interests in mind. Of relevance to this Committee, all our policy development and research is grounded in children’s lived experiences, and we work with over 300 youth ambassadors across more than 70 countries.
I will draw a little on what Damon was just saying about the patterns and the impact. I think that is what we see as well. The fact is that technology is thoroughly embedded in children’s lives now. From when they are a baby—in fact, from when they are still in the womb—a profile is being built of them online somewhere. That means that they have a different perception of offline and online from ours. We knew or grew up in a world without the internet, or with a rudimentary internet; they have had all that from birth, so they do not really discern it in that way. It is all their lives. It is not necessarily a tool or resource; it is an environment that they are in.
Something that brought that home to me recently is a project we are doing looking at the use of technology in the classroom. We were talking to a group of five and six-year-olds, and they started talking about Google and referring to Google as “he”. They told me, “He is very smart” and “We can put something in and he will tell us the answer.” I just thought that for very small children, I guess that is the kind of relationship they have with technology now.
On the impact, I think it is that the boundaries of childhood are changing online—even the age at which childhood ends. We would say that that was 18, as per their rights, but online, 13 is the age that has been set by lots of companies. That is when children have had access to this almost adult experience. For example, the first time that lots of children see pornography is by accident, and it is when they are very young. It could be violent content, or—thinking about a multiplayer game—they might be hearing swear words or adult conversations that we would deem not appropriate for children.
The final thing to say is about children’s complete lack of privacy. Children do not get to grow up in private any more. Whereas before, if you made some sort of social error, it might have stayed within your family or your school community, if they do that online now, it is there forever and they do not get to take it back. Childhood is so much about making mistakes and learning from them, but in terms of what I see happening, I can only see that becoming more and more embedded within their lives.
Q71 Freddie van Mierlo: I will resist asking too many follow-up questions, because I know that my colleagues will want to ask some. Bernadka, can you answer the same question?
Bernadka Dubicka: I am a professor of child and adolescent psychiatry at the University of York. I have also been an NHS consultant in CAMHS for over 30 years, so I have witnessed many trends over the years. In terms of my research, I principally focus on trials in depression, but I have more recently led on a sexting study in young girls, and I am involved in a digital detox study at the University of York. I published a paper with colleagues a few years ago about what clinicians should know and ask about the digital world. I am also a researcher on a unit called the Smart Data Donation Service at the University of York, which is another way of doing research in this field.
Lastly, although I am not here today on behalf of the Royal College of Psychiatrists, it is relevant to mention that I used to be the chair of the faculty there, and in January 2020 I led on our first position statement on the impact of technology on mental health. We are currently updating that, and that should be published shortly. Within that work, we have been reviewing the literature over the past five years. That is my range of expertise.
I agree with much of what has been said, both in this panel and the previous one. I am here as a clinical academic, so I will perhaps focus on the clinical aspects. Since the pandemic, we have certainly seen huge changes. We have heard about the really shocking changes in the online world that were not present when we did our first report at the college five years ago. We have not mentioned, for example, the use of conversational AI and mental health chatbots, which is another rapidly emerging field in the area of mental health, and how they can persuade people to harm themselves and so on. We are certainly seeing that impact in clinical services.
Going back to the evidence, there has unfortunately been very little done in the published research evidence base on these vulnerable populations, although there is a little bit that I could talk about. Certainly in my clinical practice, I come across really shocking examples when I ask. I am sure that you have heard many of those. Pornography has been mentioned; when I first started out in the field, I would certainly not have been treating porn addiction in young teenagers, and I have come across that on a number of occasions, as one example. Another example is a primary school child who is not sleeping at night and is self-harming, and it is only when I have asked about it that I find out it is because they are being cyber-bullied online. Parents usually have no idea about any of this.
That is another point I would like to make: this is just the tip of the iceberg. For the past decade, I have felt that this is a really important area in the field of mental health, but colleagues working in the field and professionals working with children and young people have unfortunately not been trained to think about asking about young people’s online lives.
For example, we recently audited a few CAMHS and found that, across more than 160 records, the mention of digital lives is only recorded on a handful of occasions. When it is recorded, that information does not come from probes from clinicians, but is volunteered by families. I know from our work with children and young people that they are not going to volunteer this information unless they are asked about it. I have countless examples over the years where I have been the first person to ever ask young people about their digital lives, and I have encountered some quite horrifying examples.
Of course, that is not everybody; many young people find support online and use it for positive reasons. The issue of digital equity has also been mentioned, and we know that some of our poorest young people do not have those same online advantages. The digital world is part of all our lives; if you cannot access it for education purposes or applying for jobs and so on, you are at a disadvantage. That is important to remember too.
One point I want to make is that all professionals working with children and young people, not just in mental health services but across the board, as the Southport inquiry found, should be curious and interested in young people’s digital lives, because that is their lives these days. If they do not ask, they are simply not going to tell us about it. What I see is the tip of the iceberg.
Another really important point is that digital lives are in the context of young people’s lives. We have heard about the rise in mental health problems in children and young people: the rate of probable mental health disorders is now around one in five. However, the rate of probable mental health disorder rises sharply in young women—we are talking ages 18 to 25. So, when we are talking about bans, restrictions and impacts, that is the population we are particularly concerned about and see in services frequently, particularly around girls and self-harming.
From my perspective, this is a lifespan issue—we have touched upon that already—and a wider public health issue. Although I am a child psychiatrist and mental health is obviously my priority, I feel that we have sometimes gone down rabbit holes and focused just on mental health, particularly when it comes to the evidence. But the online world has a whole range of impacts, including misinformation—we have seen that—racism and misogyny that affect all of us in some shape or form.
That can also affect children, young people and other vulnerable adults, even if they do not use social media. For example, I have come across young girls in my practice who are self-harming, not because they are using social media, but because their boyfriends are influenced by misogynists like Andrew Tate and they are engaged in a coercive, controlled relationship. The impact of that comes through in their mental health. There are a lot of issues there for consideration.
Q72 Freddie van Mierlo: Based on your experience and your expertise in depression, what is the relationship between that and digital device use and social media? Which way round does it go? Is watching social media driving people towards depression? Or is it the other way around and people who are already depressed and predisposed to that are going to social media as a means of comfort or as a coping mechanism?
Bernadka Dubicka: You know what I am going to say to answer that. That is the whole issue around whether we have enough causal evidence. As an academic, I am very interested in the evidence base. When academics have spoken about the lack of causal evidence they are talking about published quantitative data.
The problem with that is that we are light years behind the tech companies. Even when we are reviewing the data for the past five years and reading studies about Facebook, that is simply irrelevant in this day and age for young people. We are really struggling to keep up with the pace of research and the funding for research—that is another issue—necessary to find those causal effects quickly enough. We are never going to be able to catch up in terms of that.
However, the best form of research always looks at multiple sources; we call it mixed methods. We look at qualitative data, or in other words interview data. We talk to people and we use that information to supplement data points, because data points are hard to interpret without understanding what they are about.
Q73 Freddie van Mierlo: Based on your clinical practice and other sources?
Bernadka Dubicka: Based on my clinical practice, it is mixed. It is messy, because there are so many variables interacting. That is reflected in the literature. For example, I might see a depressed young woman who is an influencer. When she is feeling well, she will be giving lots of positive advice and support to young women not to self-harm. Then she will come to see me because she is feeling really depressed and has started self-harming, and she will be using the same network online as a way of encouraging her to self-harm. The effects are mixed and it is complicated.
Damon De Ionno: I agree that it is very difficult to establish anything causal about something that is so complex. What variable are you isolating? A lot of the research that has been done and is cited is about screentime, which obviously encompasses many different activities. It is also the case that if you take one child in a population off social media, they will lose contact with their social network, because all the other children are on social media.
It is true to say that when people are depressed or unhappy, they might turn to social media—it is the same thing with children who have special educational needs, for example—but that is because the rest of the world is on social media. If that social media was not there, then they might not have to turn to social media. That is a real problem that has to be acknowledged when we talk about this evidence gap, because no one can tell you what the experiment is that will decide whether social media is healthy, or whether it is the cause.
I agree with Bernadka that you have to look at the reality of children’s lives. You have to talk to parents, teachers and clinicians. You have to talk to those groups and children themselves. When you talk to those sorts of people, there is actually a very clear signal. The fact that we do not have a scientific study yet that establishes it is not a reason not to act.
Q74 Chair: Thank you very much. I will follow up with a point of clarification. Bernadka, you said that Facebook was irrelevant to children’s lives. We heard in the earlier panel about having transparency and access to social media inputs and outputs. Do you have a recommendation? Would that improve your research quality if you had better access?
Bernadka Dubicka: Yes; it is the same recommendation we made with our last royal college report in 2020. The companies need to be mandated to share their data, which is obviously easier to say than to actually do.
Q75 Chair: Why is it easier to say than to do? Is it a technical issue or a political will issue?
Bernadka Dubicka: Even if it was mandated, in this day and age it is difficult to see them sharing that data, which they would say was commercially sensitive. That is not a reason not to try, though.
Q76 Chair: Is there any definition for what data you would like to see shared? I do not want to focus on it now, because it may be technical, but I would certainly like you to write to the Committee on that, if that is possible.
Bernadka Dubicka: That is reflected in the platforms that we have up in York.
Q77 Dr Gardner: I want to aim my question primarily at you, Bernadka, but Damon, you mentioned that the type of content has changed. I am interested in looking at that and at the impact on brain development and self-regulation. In the Revealing Reality report, children reported being addicted and particularly highlighted short-form videos, infinite scroll and algorithmic content selection as key features. They described brainrot content—this very nonsensical, AI-generated, heavily layered TikTok—and also talked about brainrot in terms of mental fog or reduced cognitive abilities, which is an area of real concern for me, from spending time on useless content. In previous sessions, Professor Sarah-Jayne Blakemore explained to us about the limbic system, which involves things like emotional processing and reward processing, and how it develops relatively early and is hypersensitive in adolescence. It seems to me that social media particularly is being designed to target that very sensitive age. Bernadka, are you seeing any changes in attention, concentration or emotional regulation that you think may be linked to patterns of digital use?
Bernadka Dubicka: Within the clinic?
Dr Gardner: Yes, that you can relate.
Bernadka Dubicka: In terms of the vulnerabilities of children I see in the clinic, that again depends on the different vulnerabilities, on which we have minimal research. We know that, for example, since the pandemic emotional disorders have gone up among girls in particular, as more young girls are feeling depressed and anxious. Their self-esteem is at rock bottom. You heard in the first evidence session how adolescence is such an important time for peer approval. Particularly when you are feeling low or depressed or that you do not have any friends, that peer approval becomes even more important, and you are even more sensitive to potential cyber-bullying, for example—and we do have quite a lot of evidence on that in the literature.
Those are the sorts of impacts I see around cyber-bullying, which then impact on depression, anxiety and self-harming. The other impacts I see are from things like non-consensual sharing of nudes—we have done some research on that—which causes devastating impacts on girls, particularly when they are depressed, anxious or self-harming.
Q78 Dr Gardner: That specifically impacts their mental health, but I am looking at their cognitive functions—their ability to self-regulate and their attention, which there have been reports on. Have you seen any impact on that level?
Bernadka Dubicka: It is difficult to say at a clinical level when I see an individual child, because I do not have the measures over time to demonstrate that. But certainly, we know that there is evidence in the literature that attention is affected, particularly in children with ADHD. In terms of children with autism, we certainly see those impacts in the clinic, because they really struggle to understand the social environment online, and that can lead to lots of issues and lots of vulnerabilities online, but I cannot comment specifically about changes in cognition that I see.
Q79 Dr Gardner: Damon, you did raise types of content—
Damon De Ionno: Yes, I did that work that you were referencing.
Dr Gardner: Oh, sorry, I should have known—my apologies.
Damon De Ionno: That’s okay. We did that work and we published it in the way that we did because in my experience, when you show adults the content that children watch—if you show them a minute of many different children’s TikTok feeds—they find it a very hard watch. Most people, including leading academics, specialists in areas like addiction and clinicians, will say that it is like watching a very fast fruit machine. It does not have much story to it; it is very jarring. If I’m honest, when you hear it on the bus, it very, very jarring, and it is different from how I am sure people around in this room use social media. From just watching the content, it is hard to see that it has any benefit.
Children do not always have a counterfactual to compare their attention with, but they have been increasingly reporting in the work we do that they find it hard to watch films or TV, and they certainly find it very hard to read. Again, we cannot say that that is causal, but they are engaging in hours and hours a day of this very fast, quick-fire, broken-up content that does not have a story. They report that what is good about it is that you can just skip to the next one. You can see them doing it—I do have stuff to say about the data that social media companies hold, if you want to know, but even when you look at the data donation from TikTok that children give us, you can see that the something like 80% of videos are watched for five seconds or less. It is a very rapid activity, and the children do a scrolling motion as well when they describe it—that is the problem.
Q80 Chair: What is the data donation?
Damon De Ionno: Data donation is, I believe, due to already existing legislation. The companies offer users the ability to download the data that they hold on them. For TikTok, it is not that difficult: you can go on as a user and press a button or two, and it will download a file that you can look at. It is a set of text files, and it is not that user-friendly. The average person will find it very difficult to extract anything from it or understand what it is. However, what it does have is every video that TikTok thinks you have watched and how long you have watched it for. It provides a link to that video. It has every message you have sent, everything you have liked, everything you have bought—if you tick all those options in your data donation, then that is the data donation that is happening at York.
Bernadka Dubicka: That is the service that we provide.
Q81 Chair: That is not a donation—that is you getting your own data.
Damon De Ionno: No, the donation is from the user to us, as researchers. The donation is to us. They can donate it to us and give us the rights to process it, and then we can quite easily—because we have more skills than they do, since they are 12 or 13—create dashboards and understand, for example, the average view time on a video.
In terms of your question on what the companies could do, they could make that data much more accessible than it is. They could also do analysis on it and tell us how many young people are watching into the night—and they do know, because we can see it from their data. When we analyse their data, we can see that something like 20% of what children of 13, 14 or 15 are looking at on TikTok is happening after midnight.
Q82 Dr Gardner: There has been significant reporting on the addictive nature of this at an important time of development. Do we have evidence about the impact on cognitive abilities? Have you seen it specifically in your clinic? Is there academic evidence to prove it?
Bernadka Dubicka: It is difficult to give anecdotal evidence. My colleague Lisa Henderson did a small pilot smartphone detox study in schools, with a small number of children and young people, which showed that over a couple of weeks of a smartphone detox their mood, attention and sleep improved. That study is now being scaled up to a large randomised controlled trial in York, so there is some evidence.
The other thing I should have mentioned in the clinic is that, over the past few years, there is a pandemic of sleep difficulties in children and young people; they always have their smartphones by their bedsides, and it is extremely difficult to get them to remove them. That has huge impacts on their education.
Dr Gardner: My 17-month-old grandson does the scrolling motion, and that is what he wants to do on the phone. He does not even watch what is coming up—it is that motion that he wants to do. I find it very interesting.
Chair: But he does not have a phone, surely?
Dr Gardner: No, no, but if he can get hold of one he does that. My daughter tried to restrict him, but he is drawn to this thing, and he will do that scrolling gesture all the time. It is very odd. He is not odd, but that is odd.
Q83 Kit Malthouse: We are obviously a Committee in search of some kind of certainty, trying hard to feel our way to conclusions and recommendations. What we have found from quite a lot of witnesses—with some exceptions—is a sort of equivocation in the face of even their own direct experience. It is all, “Well, you know, on the one hand, on the other hand, actually overall the whole thing’s beneficial,” and all that kind of stuff. I am interested in what you would say, for certain, about the impact of the current digital environment, particularly on kids and adolescents. In your experience, are there any emerging certainties that we need to have in our minds?
Colette, shall we start with you? You are at the frontline: you see it, you feel it and you know. You are the expert. The evidence might be taking time to catch up, but where is the certainty in your life?
Colette Collins-Walsh: I will answer something else first and then come back to your point. You are absolutely right that the retort, certainly from the tech industry itself, is often, “You can’t prove this. You cannot prove that this was us. There are lots of other things.” What we do know very well, though, are children. We know what is good for them. Even in your own observations of, say, a very young child, you know that they need lots of eye contact, lots of physical affection and someone to talk to them a lot. That is vital for their development, so if they have a screen in front of them, for however much time, we know that that is not right. As a society and through science, we have decided that we do not think that is right. The certainty is about what we know is good for children.
A certainty in terms of the impact is that we know that this is hugely damaging to their privacy—indeed, all our privacy—just through use of data. But children do not know it is happening. They do not understand. They are much more vulnerable to the negative impacts of having their privacy infringed upon constantly, every single day. I should also say that this is not just social media. This is throughout their lives. This is in the classroom, through edtech.
Bernadka Dubicka: In terms of the certainty, you are not going to find that from academic papers, but we need to look at this as a wider picture—as a societal issue and a mental health issue. For me as a clinician, there is this double standard: we have lots of pathways and procedures we should follow for offline safeguarding, but when it comes to online safeguarding, that just seems to go out of the window. We do not wait for a randomised controlled trial of the sexual grooming of children in the offline world, so why are we waiting for this?
There are multiple examples of harms caused. It is not just around mental health; I would urge you to broaden the discussion to, for example, cyber-crime and misinformation.
Q84 Kit Malthouse: That is interesting. The origin of much of the concern is parents and parents’ groups. That is where it has come from. I am interested that that your profession—presumably lots of your colleagues are experiencing the same thing—has not been ringing the alarm bells for even longer, saying, “Oh my God, you would not believe it! Emergency, emergency!”
Bernadka Dubicka: There are several reasons for that. It is not part of training for most professions, so there is a massive gap in terms of education for professionals. If people do ask about it, what are they supposed to do? Refer to social services? Social services are not trained. They are overloaded. No one is going to take this stuff down. We know that the tech companies do not act on it. When people find these sorts of harms online, there is no proportional response that they can ask for.
It is the fear of not knowing what to do with that information. It is too big of a thing to address. Those are some of the reasons that my colleagues do not ask. But interestingly, when I do my usual talk—I did it for psychiatrists recently—all the questions come from them as parents: “What can I do with my eight-year-old?” As parents, they are extremely concerned.
Q85 Kit Malthouse: The royal societies and the associations are pretty voluble about other developments, but they are never voluble about this. I do not understand why.
Bernadka Dubicka: The corporate playbook is also at play, isn’t it? We have seen that before with gambling and tobacco. They know how to play this game. They know how to shoot down potential concerns and potential evidence. It can be quite a hostile environment to speak out in as well.
Q86 Kit Malthouse: Damon, where is the certainty in your life?
Damon De Ionno: Before I answer that, it is a very private thing. It is different. If it was happening on a screen in the living room and everyone could see what was happening, maybe we would be talking about it in a different way. The fact is that everybody is in their own world—when you are on your devices or you are on social media, you are in your own world; you do not see someone else’s experience.
This goes back to the question that was asked at the beginning. Why would clinical psychologists and psychiatrists think about that? They do not experience it—they do not see it. Why would they ask the question about it? It is not going to appear in the literature for a long time, for many of the reasons that Bernadka has already given.
Q87 Kit Malthouse: You would expect that if large numbers of young people start showing up in psychiatric clinics across the country—
Damon De Ionno: They do.
Kit Malthouse: If it was something physical—for example, 10, 12 or 20 years ago, large numbers of young girls were showing up in central London and other areas with signs of FGM. Suddenly there was an enormous fuss. We had legislation to deal with it. It was picked up by the clinicians. They were all saying, “Oh my God! You won’t believe the phenomenon that we’re seeing.” We have not seen that with this.
Damon De Ionno: Well, we have in various subject areas. Some of them are quite contentious, but we absolutely have seen that. The trouble is that in academia, for good reason, attribution is very cautiously applied. We are talking about very complex phenomena which shape our society and our culture. They act in very, very subtle ways, as well as directly. That is why it is very difficult to pin down. That is why when we are asking for that causal effect, it is very, very difficult to get that attribution. It would not go well for clinicians to start saying, “I think it’s definitely this,” because there are standards in terms of evidence.
Bernadka Dubicka: It is not that we do not see it; it is just that we do not ask about it. So many young people are not going to volunteer this, particularly if they are doing things that are illicit or things they should not be doing, and also, it is shameful—if they have been sharing nudes, they are not going to share that willingly.
Q88 Kit Malthouse: You are presenting us with an impossible loop: we cannot research because we cannot research, so we can never conclude.
Bernadka Dubicka: We need to ask.
Q89 Kit Malthouse: Would it be fair for the Committee to conclude that it is clear that the development in the digital life of children has interfered with their hitherto normal developmental processes on a widespread basis?
Damon De Ionno: This is how I was going to answer the question you asked about what I am certain about. We are certain that we see children experiencing harm. Those harms have happened—grooming, sexual abuse, eating disorders, being exposed to pornography, effects on sexuality. They can describe those things themselves, and we can see the evidence for those things.
In terms of the stuff around cognition and their development, that is going to be very difficult to establish, but what we can say is that there are large numbers of children who are spending a lot of time doing some activities for which we cannot see any benefit. If they were spending that amount of time at school and we did not see any benefit, and we saw some harm and they were reporting some distress about it, that would be a serious issue.
The displacement is what we can see—it is the lack of other activities. Critically, if you spend time on a screen, on social media, scrolling TikTok for hours and hours on end rather than having some form of connection, you are not talking to a human being. You are not interacting with a human being. You are not learning how to interact with human beings. You are not developing that emotional machinery. You are not developing the skills, and we do see that play out—we see massive rises in social anxiety. Is that causal? I do not know, but they are not doing some of the things that they need to do to not be anxious.
Q90 Kit Malthouse: So we would be right to conclude that it is interfering negatively with the development of the adolescent brain.
Damon De Ionno: It is stopping them from doing other things that might be positive, yes.
Q91 Kit Malthouse: That is not quite the same question.
I have a second question, particularly for you, Damon, because you are on the frontline of this. What do the kids say about this? As you said earlier to Allison, we have seen the BBC documentary where they persuaded a whole year to remove their smartphones, and they all said, “Oh God, what a relief! Finally, we’re free of the tyranny of Snapchat.” What do they actually say?
Damon De Ionno: They want to be able to message each other—that is not going to change. But about two years ago, there was a shift in what children said around short-form content. Prior to that, it was almost impossible to get a child to talk about the negatives of social media. They would say how brilliant it was, how positive it was and how much they were learning.
About two years ago, across demographics, children started to speak differently, and that is when the term “brain rot” became popular. It came from children; it did not come from us. The narrative shifted from children to say—and they also talk about other children and younger children—“They probably shouldn’t have that.” That is when they started saying that. I do think the content has changed, and that the separation between the social side and content consumption is part of this. They started to say that it was a waste of their time; they did not feel so great afterwards; they knew that they were spending more time on it than they wanted to; they knew it was interfering with other things. They still will say, “I don’t want to be off it if everyone else is on it,” and that goes back to the idea of missing out, but they have started to say, “If no one had it, that would be better.”
So you are asking the wrong question, because if you ask the child, “Do you want us to ban social media?”, they think, “I won’t have access to all the things that I have, to the people I know or to the content.” They don’t think, “Well, no one else will have it either, so we will probably spend more time together.” I have sat with groups of boys, for example, who say, “Well, of course I like playing football in the park more than playing on FIFA, but if I went to the park, I’d be the only one there.” That is the challenge.
Q92 Kit Malthouse: It is a collective action problem, basically.
Damon De Ionno: Yes, basically.
Q93 Kit Malthouse: And on that basis, even if a ban is imperfect, it might at least get, I don’t know, half the kids back out in the park.
Damon De Ionno: Yes, it might. I think, realistically, children and parents do want to spend less time online.
Chair: Thank you, Kit. I think we are going to talk more about displacement activities with Daniel now.
Q94 Daniel Zeichner: We are. Good morning, everybody. I think we may have covered quite a lot of the displacement activity already, but I want to go back to something that Colette said earlier about nothing being forgotten, and also the data donation issue. If all that data is being stored on what young people are doing, and it is not forgotten, is there not a danger in future that, when they come to apply for jobs, university and the rest of it, people could look back and draw patterns? You could probably fairly accurately predict that their outcomes are going to be worse if they have spent hours watching particular kinds of content. Do you see what I am getting at here? Is there a danger, and what do we do about it?
Colette Collins-Walsh: Yes, absolutely there is a danger, although something I found quite interesting recently in Revealing Reality research—Damon can correct me if this is the wrong interpretation—is that there is a bit of a trend of children, or certainly adolescents, being much more aware about sharing stuff online. They think it is a bit cringe to share so much online. But to your point, yes, that is a problem.
Frankly, these companies should not really be storing children’s data under GDPR and the age-appropriate design code. They certainly should not be using it in such a way that could harm them; they should only be processing it in those children’s best interests. In terms of what we can do, I think we could enforce data protection properly in this country. That could be a good start.
Q95 Daniel Zeichner: Very good. On the displacement point, Damon, we have covered quite a lot of that, but do you get any sense that children understand the trade-offs that they are making, or are they not in a position to make those decisions?
Damon De Ionno: Yes, and they are often increasingly articulate about it. They can identify specific features and elements of the apps and the way that they work. They are very aware that they do not intend for their usage to be what it is, and they talk about wanting to reduce it. They talk about it in the way that someone might talk about a gambling habit that they felt was a bit out of control: “I start and then I can’t stop.” So they are quite aware.
On the data donation, one of the issues is that many children are registered as adults on the platforms. Because they want access to features that are age restricted, they register as someone older than they are, or sometimes their parents use their own date of birth because they see it as a way of not giving out the child’s personal data. That is a really serious issue. I was listening to the previous panel, and yes, the platforms have a fair idea of how old those people are, even if their stated age is different. They advertise to them differently, so they have some idea. But data donation is a big issue, because a lot of children are registered as adults.
Q96 Daniel Zeichner: In terms of displacement, whether it is content or communication, the time is still being taken from something else. Should we actually just be saying that there should not be so much time spent on screens in general?
Bernadka Dubicka: Yes. Can I mention two things in relation to that? You mentioned adolescents, but you have not talked about very young children. I was really pleased to see that the Government have produced some guidance around them, but we do need support for parents, such as health visitors to help educate parents. The World Health Organisation published a report in 2019 looking at the evidence base around the impact on development of very young children. It concluded, even then, that too much screen time had significant detrimental impacts on development. We have had more longitudinal studies since then. That is important to mention.
I was thinking about the second issue of displacement of activity before I came here. Arguably, society has partially driven children and young people on to social media because of the lack of external activities. I have seen that over time. I am a medic and prescribe medication at times. I can put people on waiting lists for therapy but, by definition, it is a long list.
We do not offer children and young people very much. Sure Start programmes shut down significantly. I would like to be able to prescribe outside activities, sports, culture, music or whatever for the young people I see in clinics, but access to those activities is difficult and limited. There are no barriers to accessing the internet. As long as you have access to the internet and a digital device, the world is out there for you.
That is not the same for offline activities. If we are talking about bans and displacement, we also need to give young people alternatives, including mental health support. As you know, we have huge waiting lists. If they cannot access things offline, they will turn to mental health chatbots.
Q97 Daniel Zeichner: I have one final observation. I worry about relying on educating parents. I visit schools every Friday; I am afraid the lesson I have learned from teachers is that the problem is quite often the parents, not the children.
Bernadka Dubicka: That is the issue: modelling.
Q98 Daniel Zeichner: Frankly, we know it will be some of the most vulnerable chaotic households where the children will be most vulnerable.
Bernadka Dubicka: That is why it is a societal issue. I agree we need to intervene early and think about how parents are using these devices.
Chair: Martin, did you want to come in here?
Q99 Martin Wrigley: I want to pick up on your societal issue. I am curious, because we talk about this as if it were new and different and never happened before. I don’t believe that to be true. Technology has made things faster and more available. Fundamentally, do you think that social media is any different from profitable, harmful and addictive products that we have seen before, such as tobacco and gambling?
Bernadka Dubicka: It is obviously not the same. My concern is as a medic prescribing medication. There is regulation around the industry and companies have to prove safety before they are released on the public. That does not happen here. The financial and taxpayers’ onus is for academics and the public to demonstrate harms. That seems to me the wrong way round.
Colette Collins-Walsh: Parking the cigarettes and alcohol question, you sometimes hear, “They thought this about television and games. Is this the same?”
Chair: Reading.
Colette Collins-Walsh: Yes. This is different because, whereas a child might have chosen to pick up their phone in the first instance, or go on a games console, they made that choice, but the choice to stay on there is not theirs. The way the platforms are designed infringes on their agency to put it down. This is something we hear a lot from children. They say, “I feel addicted or wish I could put my phone down.” Or they reflect, “I went to bed last night, but I could see there was a notification on my phone, so I picked it up and went back into my phone.” There is vulnerability as children in their agency, which is different from watching TV, reading or gaming.
Q100 Martin Wrigley: But no different from smoking?
Colette Collins-Walsh: I would go back to the comment that those products have to come forward with answers to: “Are they safe?” and “Who are they safe for?” We have details on a cigarette packet to say who they are for, but we do not have that for social media.
Q101 Martin Wrigley: So it should be defined as a product and expected to comply with product safety rules.
Colette Collins-Walsh: I would argue so, yes. If I could go back quickly to the point made about parents, I would say this in their defence. Last Monday, when the social media ban was announced, was probably the first time the Government made a big statement that, in their view, it is not safe for this age group, whether or not you agree. If parents are drawing on their own experience, they do not have experience of these technologies. How are they teaching their children? I would defend parents a little in that situation.
Chair: That is a very good point. Thank you. We now come to Maya.
Q102 Maya Ellis: I am going to build on the conversations we were just having and talk about the big picture a bit. I mentioned a couple of things in the previous panel’s session. First, we know that youth services in this country have been cut by 70% on average over the last 10 to 15 years, so we have had decimation of that. I am also really interested in the push and pull of all this. There is the addiction pull of social media but, as we have touched on briefly, what else is driving people to that?
Damon, you asked, “Why are we doing this when there is no benefit?” This is very big picture, but I hope you will follow my train of thought. There is one benefit—a safety benefit, in the sense that children can stay in the house when parents are working. We know that there are more dual career households now and that parents are working longer hours. In the economic calculation balance, society is getting much easier handling of children if we stick them in front of a screen or social media.
I have pushed the Minister for Early Years on this. When we announced the guidance on screen time for early years, the one thing I pushed back on—I would be interested if there were any academic evidence on this—is that we are not properly looking at screen time as childcare. We can put all these restrictions on screen time for young children, but I would hazard a guess—this is anecdotal—that when most parents have a child, they do not think that the exceptional way to raise that child is to put them in front of a TV for an hour or two. They do, however, come home at half 5 having picked them up from nursery, and then need to cook tea and tidy the house, so they need a bit of time for that child to be quiet.
I am worried that we are not considering those massive push factors for parents. Again, I will disagree with my hon. Friend the Member for Cambridge on challenging parents. I do not think it is about educating parents. I think it is about supporting parents, because I am not sure we have a society now that can cope with the interaction needed. Colette, you were talking earlier about this. We know the things that babies need in terms of eye contact and talking to them, but the majority of parents that I speak to often do not have the capacity to do that, because of wider shifts in society putting those pressures on them. I am interested in those thoughts. It is a broader question: do we think that digital devices are filling a gap created by society? I am interested in whether there is evidence on this and what your thoughts are on that.
Damon De Ionno: We do a lot of work with parents. There is some work from the ICO that will be published in a month or so about how parents and carers manage technology in the home. Parents are trying harder and harder. They are generally quite aware of the problems.
As a whole, there is massive anxiety from parents about children outside the home. That is undoubtedly the case. Tracking is ubiquitous. There is some fear creation around that in the tech industry, which is a separate, bigger problem about surveillance and basically allowing children to grow up. There is comparatively less worry about some of the dangers online. There is even less worry about the potential cognitive impacts or lack of development of other skills or displacement, which we were talking about earlier, but people are aware of it.
I hate myself for saying this, but parents could put their child in front of the TV instead. I would have said that that is not a great thing to do, but given the choice between TikTok and children’s TV, children’s TV does contain stories. It is regulated in some way. There are standards. There are professionals doing it for the child. That is not the case with social media content. Social media content is produced to make money and whatever sticks, sticks.
There are alternatives. There are also children of all demographics whose parents do not give them social media or technology and they do thrive. I can think of many examples, but even when those children finally get the technology, their other activities tend to decline a bit, even though they have access to them. It is complicated but, overall, I do not think that there is just no option for parents. Once they have given it to a child, however, or once a child has experienced it somewhere else in some other part of life such as a friend’s house, it is very difficult to go back.
There is a quote from that upcoming report. A lot of parents now, who are not so middle class and not so much fans of the anxious generation, are saying things like, “I wish I had given my children a device a little bit later.” It is really hard to go back. You cannot put the toothpaste back in the tube. What I would say about any legislation is that many parents use the age of 13, even though it is not the law. They say, “Social media is not until you are 13”, as a way of telling their children of 12 that they have to wait, whether that is legal or not. I think that if an older age was there in that case, many parents would be empowered to say to their children, “You have to wait.”
Colette Collins-Walsh: I agree with Damon. I defer to Revealing Reality’s work on parents, because we work mostly with children. I would add, however, that all screen time is not created equal. I think not in your previous session, but a couple of weeks ago, you talked about facetiming with grandparents. For a young child, that is perfectly acceptable. How else would they see them? This is about not so much the screen necessarily, but what is on it. If it is social media, we know that is bad; however, my nephews watch “Peppa Pig”, but on a tablet and not for long. All screen use is not created equal. They use screens in schools too, so we might say they had an educational benefit, perhaps. That is all I would add.
Bernadka Dubicka: I agree, but would add something about the different types of screen content. Of course, very young children have AI toys now, and there is developing research and concerns about that as well. Then you have the interaction between parents, who are on their phones and screens, and children, so the TV set that used to be in the living room is now out there 24/7. Parents are also on screens, and it is that joint interaction and what they are doing online, how their data is being accessed and who is contacting them.
Q103 Chair: The evidence that we have seen showed a big difference between content that was engaging and two way, and content that was one way—so content that is just being consumed and content that requires a response. Collette, do you agree with that?
Colette Collins-Walsh: Will you repeat the question?
Chair: Some of the research offered to the Committee suggested a big difference in impact between one-way content, which is just being consumed, even if you scroll through, and content that requires a response or some kind of engagement.
Colette Collins-Walsh: Yes. From the perspective of children—we touched on a little of this earlier—I guess it is the difference from passive content. The young people we speak to are quite discerning about what is good and bad time spent online, even if they spend the whole day using different devices. Last summer, we asked one of our youth ambassadors to keep a tech diary for a day, and he spent the whole day using tech. It was Saturday, but he watched YouTube with his mum in the morning and binged an entire Netflix series in the evening, but when he made a reflection at the end of the day, the one thing that made him feel a bit rubbish was that he had spent half an hour watching YouTube shorts, which is passive. The other things, I guess, were more engaging—he was also playing a game. The things that were engaging were not so bad, but he himself was reflective and regretful about that bit of time he spent because it was so passive.
Chair: I know we are running over time, but we have an important subject to consider with Freddie.
Q104 Freddie van Mierlo: I just have a quick question. Damon, you differentiated passive content consumption and messaging in particular but, Bernadka, you highlighted some of the dangers of messaging, the sharing of non-consensual nudes, cyber-bullying and some of the impacts you have seen in your clinic. How do we get that balance right? The banning of social media, or what the Government have proposed, is primarily about passive content consumption, and they will allow messaging apps like WhatsApp, Signal and so on continue to be used, so we would not eliminate those dangers through that ban. What do you think about the tension there? Are further restrictions required—perhaps on features, WhatsApp and that kind of thing—that may help prevent some of those negative harms?
Bernadka Dubicka: I welcome the conversation and the fact that the Government have taken a stance—as has been said—on the fact that there are dangers. That is an important and powerful message. But yes, there are concerns about a social media ban, because as we have heard repeatedly, it is about the design of the products and the platforms, and which platforms have those unsafe designs. It is more about the design features that need to be tackled—we have touched on that already.
Of course, it is not just social media. You can connect with strangers in various games and on platforms like Discord, so you need to think about the wider digital ecosystem, because if you ban just one or two things, kids will be on something else.
Of course, there is the right of children and young people to access important content such as education, and to make important connections such as with friends and family. There is a very difficult balance to tread, but we do need to look at the safety of the products themselves. That is more your area, isn’t it, Colette?
Colette Collins-Walsh: I entirely agree. We agree with the Government’s assessment that the current deal on social media for children is unacceptable, and we support an emergency brake on that, but that cannot be the end of the story. We were concerned to see that the Government’s progress update said that they had a lot of concern about addictive features, but they felt they had addressed that through the ban, but as we know it is pretty limited, and a lot of those design logics are everywhere. It is in AI chatbots, gaming and sometimes even edtech, so it is almost not going far enough. We support an emergency brake because it is so harmful, but that cannot be the end of the story. We need to see much more to deal with the business models and the underlying drivers of risk to children.
Damon De Ionno: I have been making a separation, and I want to make a further separation. There is messaging and communication between people who already know each other, and then there is everybody in the world, who may or may not wish to talk to a child. That is of concern. If we were to draw an offline parallel, we would not think that was acceptable at all.
Q105 Freddie van Mierlo: Do you think there should be unlimited access to messaging each other?
Damon De Ionno: No, definitely not. I have heard about far, far too many children having bad experiences. Although you will hear people say, “I have made a friend online”—
Q106 Freddie van Mierlo: Specifically for children who already know each other. Should they be allowed to message each other all the time, no matter what?
Damon De Ionno: You are going to find it very hard to stop that, and parents want their children to be messageable all the time, too. I don’t have a strong view on whether it should be limited—maybe at night, for example. The key distinction is whether they are contacting people they don’t know. I think everyone would agree that we do not want seven-year-olds meeting random people in the world. That is a separate issue.
In terms of the features and applications, we have mapped it out and, if you do a timeline, there is huge convergence. Basically, every social media platform and application adopt the things that work on other applications. While TikTok is used predominantly for short-form video viewing, it has messaging, streaming and all these other functions, such as streaks. It has all the things that the other things have. Snapchat is predominantly for messaging, but it has short-form video content in the way that TikTok does. Instagram obviously introduced reels, which was copied from TikTok. You definitely need to be thinking about the features; it is not a case of individual apps. If you do it to individual apps, it will just be displaced.
Q107 Chair: Great. Thank you very much. We are nearly out of time, but I want to raise the key issue of inequality. You have talked about how social media and screen time can differently impact different demographics and different backgrounds. There is particularly the impact on children with neurodevelopment disorders such as autism or ADHD. Bernadka, can you briefly summarise which groups are most vulnerable?
Bernadka Dubicka: Again, there has not been much research on vulnerable groups, but I am aware of one study that looked at almost 3,000 young people and found that the more offline vulnerabilities a young person has, the more they are exposed to online risk. There is some evidence to show that from the literature too.
Obviously, I work in CAMHS, so from my personal experience, it is young people with autism. Girls who have emotional difficulties are disproportionately affected and seem to be much more vulnerable, particularly around self-harm and eating disorder content. We are increasingly seeing that in boys too, with the whole looksmaxxing trend; we are seeing more and more boys who are susceptible to body image issues. Also, there is misinformation on diet for girls—and for boys now—in the obsession with supplements, proteins and so on. If they are already vulnerable by virtue of being autistic, or having depression or anxiety, they can be much more sensitive to that. Other groups, such as looked after children, can be very vulnerable, particularly to exploitation online. It is children who are marginalised and excluded, including through poverty.
Q108 Chair: That vulnerability carries over.
Bernadka Dubicka: Yes. Of course, the LGBT community struggle to find safe spaces offline, so spaces online can often be quite helpful for them.
Q109 Chair: I will finish with a question that is for each of you. We know that the Government are looking at a social media ban and other measures to address the friction in social media. Could you each tell me one thing, in addition to the Government’s proposals, that you would like us as policy makers to do to address the impact of digital devices on young people? We have also heard that the Government are too slow to react; technology is moving too quickly. Could you give us one principle that we can adopt to help us to address the future evolution of the technological impacts that already exist?
Damon De Ionno: At a small scale, we try to show the reality of—
Chair: Just one thing.
Damon De Ionno: I would like the platforms to show the data on how they are being used and make it public. In terms of principles, they have to be liable for outcomes.
Chair: That is great. Colette?
Colette Collins-Walsh: End the commercial profiling of children—just stop it. So much of what we are talking about is through commercial profiling. On principles, I return to what I said earlier about remembering what we do know about children and not worrying about the causality—notwithstanding this Committee’s excellent work to try to get to the bottom of that.
Chair: Thank you. Bernadka?
Bernadka Dubicka: Mandate safety by design and the safety of design features, and get the platforms to pay for all the issues that have occurred and what we need to do through public health.
Chair: And the principle?
Bernadka Dubicka: The principle is safety by design, which should be mandated.
Chair: So that is both a principle and a specific requirement. Excellent. Thank you all very much for your contributions; it has been a fascinating discussion. We have gone over time for this session, but as you can see, the Committee has a huge interest in this.