Oral evidence - Children and the internet

Simon Milner and Katie O'Donovan.

Q108 The Chairman: I rise to welcome you, Katie O’Donovan and Simon Milner. Thank you very much for joining us. We are very appreciative of you giving up time to be with us. Our inquiry is concerned with children and the internet. It is a big subject for us. You are very much in our thoughts as we get deep into this world. Could I ask you if you would be very kind and introduce yourselves, say where you come from, and make any opening remarks you would like to make? Katie, perhaps we will start with you.

Katie O'Donovan: Thank you for having me here today. It is good to join you all and to follow your inquiry as it has been going. My name is Katie O’Donovan. I work for Google UK in the public policy team, and I am responsible for our child safety work. Prior to that I worked at Mumsnet, a parenting website, where I was responsible for our policy and campaigning work there. Currently, I sit on the board of the Internet Watch Foundation and am a member of the governance board of UKCISS on online child safety.

Simon Milner: My name is Simon Milner. I am a policy director for Facebook, based in the UK but covering a number of countries, including the Middle East and Africa. I have similar credentials, in the sense of having sat on the UKCISS board and on IWF, and I am very much looking forward to our discussion today.

The Chairman: Thank you both very much.

Baroness Benjamin: I am sure both of you agree that in order for us to keep our children safe, protected and informed we all need to act and play our part in a responsible way with integrity. What do you see as the role of your platform in helping to safeguard children online and inform children, parents, schools and others about the safe usage of your services?

Katie O'Donovan: We take our responsibility very seriously. Google as a company is 18 years old, which in internet terms is quite old. That gives us the ability to evolve as a company and develop our understanding of the significance of this issue. It is a matter we take very seriously. It is probably helpful to describe the approach we take in three ways. We have very strict terms and conditions for our platforms and where we host content. We ensure that those conditions are adhered to, so where people notify us that those conditions are not being met we will review those and, if necessary, take action.

We also believe that technology is really important. As a technology company, we feel it is within our gift to help on this issue. For example, we have tools like SafeSearch and restricted mode search for YouTube, which parents can lock on so there is a safer environment for their children in relation to those products.

We also have flagging mechanisms on YouTube where we can use technology to help us to respond quickly. We are also able to build bespoke products. Last year, we built and launched a product called YouTube for Kids. That is a safer and more collated environment for kids to enjoy and experience YouTube.

The final part of our approach is to recognise that technology cannot be the only answer. Some of these are societal issues that we have all been grappling with for many years and generations. Certainly, the internet makes them different and can complicate issues, but often they are not uniquely a technological issue. Therefore, part of our response is also to invest and work in partnership on education to help young people develop skills, resilience and intelligence and to be empowered so they can act more safely online. We ourselves run projects like Internet Legends, but we also work in partnership with many other safety organisations in the UK.

Simon Milner: Facebook is a bit younger than Google. We are just over 12 years old. I have been at the company for almost five years. During that time, I have probably spent more time on the issue of safety, particularly associated with young people, than on any other issue. To me, that is a demonstration that nothing is more important to us than the safety of people on Facebook, particularly vulnerable groups like the under-18s.

Like Google, we have a multifaceted approach, including what our policies are, the employment of hundreds of safety experts, our partnerships with safety organisation, which I am happy to talk about more, and with law enforcement. Some things go beyond safety organisations and need to involve law enforcement. Above all, it is about constant learning. In what I am talking to you about today I will demonstrate just how much the company has progressed over the past five years in what we have learned as to the best role we can play in helping to keep young people safe.

We are very much part of a safety chain, if you like, which involves young people themselves—there is an awful lot that young people do to help one another—their teachers, parents, other technology organisations, from whom we learn as well, safety organisations, academics and government. A whole group of us are involved in doing this, and we absolutely recognise that we have a central responsibility, given that this is about the use of our platforms.

Baroness Benjamin: How do you get to them?

Simon Milner: How do we get to whom?

Baroness Benjamin: To the people you want to know about your services and what is out there for them.

Simon Milner: It is a combination of things. It is principally through the service itself. When people first sign up to Facebook there are a number of steps to go through. We also make sure there are regular reminders about things like the privacy settings people have on their accounts, because we want to ensure that if people are sharing publicly they are reminded of that and understand what it means.

Baroness Benjamin: How are they reminded?

Simon Milner: They are reminded through a little pop-up that may come up. When people first join Facebook their default setting is friends only. This is for everyone, those under as well as over 18. If somebody decides to change that and wants to share publicly, we will notice that. Maybe after two or three times we may say, “Do you know you are posting publicly? Are you sure you want to be posting publicly?” We do that through the service itself and our help centre. We have just relaunched our safety centre as well, which is focused particularly on families.

There is also the very important role safety organisations play, because they reach directly to young people in their schools and can directly interact with parents and teachers. We do not rely just on communicating online; we communicate offline, often via these safety partners.

Baroness Benjamin: Do you have measures to ensure that children cannot search for certain items, such as suicide, self-harm, pornography and so on? What happens if children do come across that content, whether it is inadvertent or deliberate?

Simon Milner: Facebook does have a search feature, but we are not a search engine for the wider internet. I am sure Katie can talk about the role of search and Google in that. When it comes to the issues you have raised, particularly around suicide and self-harm, our focus is much more on how we can ensure that, if somebody is showing signs of such distress on Facebook, there are very straightforward ways they can get help through Facebook, or that their friends can get help for them. To give you an example, if someone you know is showing signs of distress, at worst is saying, “I am about to kill myself”, or, “I am going to take these pills”, there are very good tools on Facebook for you to get help for that person. If it is really serious and you alert us to it, we can contact the local police, who can physically help that person in distress. More often, it is a case of intervening in their Facebook experience and saying, “Hey, somebody you know is worried about you. Here is where you can get help”. In the UK, that would mean giving them details of the Samaritans so they can reach out and get that kind of help for themselves. That is the area we focus on. It is not about search for us; it is much more around when somebody is showing signs of distress.

Baroness Benjamin: What sort of response do you get if you tell somebody you are concerned about their behaviour and you give them guidance? Are they susceptible to this? Do they appreciate the fact that you have done so, or do they think you are interfering?

Simon Milner: Typically, we get a positive response. The best experts in this are the suicide prevention agencies. They have told us that they find this an incredibly valuable way to reach people at the most acute times of distress. That is the hardest thing. When you are walking down the street and see an advert for the Samaritans you may be feeling fine. It is only in those dark moments when you will not be and you do not necessarily see that advert. Because we can provide that content to you directly at that time, often, it can be exactly when you need it.

Q109 Baroness Benjamin: What policies do you have, Katie O’Donovan?

Katie O'Donovan: We have a number of different measures. Google as a search engine absolutely seeks to deliver the world’s information to everyone and make it universally accessible. That is a wonderful opportunity for so much of the information, but, as you rightly pointed out, there are certain areas that can be quite difficult and challenging, particularly where content is not illegal but might not be appropriate for everyone, for example those who are more vulnerable. SafeSearch, which is a product we developed, can be turned on for any Google user. It can be locked on and password-protected, so a parent can do that for a child. That means it delivers only safe research results, so, particularly for pornographic images, it will restrict the corpus of the search results that you see. An Ofcom study released last week showed about 50% knowledge of that among parents. Therefore, it has quite a high knowledge base, but we could absolutely do more on that. It is of real interest to us to think about that and work with partners on it.

We also have a system called autocomplete. Often, when typing something into Google we want to make it easier for you to get the information you want faster. For example, if the question is about what time a film is being shown at a cinema, we can also complete that for you, but we realise there are particular topics that are very sensitive and you do not want autocomplete for that. We do not do autocomplete for terms to do with suicide or swearing, for people seeking pornographic material and for people seeking extremely violent material. People are still free to type those queries into the search engine, but it gives a bit of friction; it does not make it as easy as it is to find other things.

We work very closely with organisations on some of the controversial search terms. If people are looking for information around suicide, we have worked with the Samaritans to develop what we call OneBox. That is a noticeable block that comes up at the top of your page. If you search in the UK for terms around there, you will be pointed straight to the Samaritans and their helpline number. The Samaritans have told us that that works very well for them. They often get phone calls where people have seen that number, and they are able to provide the expertise and support they do. Similarly, we do that for people who are looking for online child sex abuse material. We work with the Lucy Faithfull Foundation and the Stop it Now Coalition. They present information that is relevant to those people. We have a wider grants programme to enable charities to advertise on all our search terms for this.

Baroness Benjamin: Do you ever take anything down?

Katie O'Donovan: Google is a search engine. We do not host the sites that are linked through. Those are hosted by private companies, individuals or charities themselves. It is difficult for us. We do not have the ability to take down content from the internet, but we can de-list content that is notified to us as being illegal. If there is illegal material in the UK or in different jurisdictions, we will not link through to that once we are made aware of that information.

Baroness Kidron: Is there a public list of things where you do not autocomplete?

Katie O'Donovan: Our policy is public. I am not sure whether it is an exhaustive list.

Baroness Kidron: We could see if it included radicalisation terms or pro‑ana sites.

Katie O'Donovan: Yes.

Baroness Kidron: Therefore, we could get an idea of the big picture.

Katie O'Donovan: Yes. Radicalisation sites is one of the issues in there, and I can certainly give you details.

Baroness Kidron: We would be really interested to see what that list looks like.

Katie O'Donovan: We have no problem in doing that.

Q110 Lord Sheikh: With regard to self-harm and suicide, I am quite encouraged by what both of you have said, particularly working in conjunction with the Samaritans, the police and the authorities. How is it monitored? For example, if somebody puts an item on Facebook or Google and is in a state of distress at that moment, quite often, that is the opportunity, when that person is down. How do you monitor that? How do you keep that under review to make sure that we render help at the right time?

Simon Milner: We do not monitor what people are doing on Facebook. However, when somebody shares, they are doing it for a reason. Typically, they are not sharing so that anybody on Facebook can see it; they are sharing so their friends can see it. I expect we are going to talk a bit later about the internet being always on, but that is one of the advantages. If you are in distress, usually one of your friends is awake and looking at their Facebook news feed at that time and will see it. That is what people are doing; they are reaching out to their friends. The key thing we want to ensure is that when that happens, if their friend spots it, there is readily accessible help.

Lord Sheikh: Could the friend get in touch with you, for example?

Simon Milner: Yes; they can report it to Facebook. We can give them some language to communicate with their friend. We will say they might want to send a message to their friend saying such and such, which will include information about phoning the Samaritans. Therefore, we can help. Typically, these are young people; it is peer to peer. We want to ensure that, if a young person sees another friend in distress, there is material readily available for them to be able to help. We know that this works. In extreme cases, we also know that when they report it to us we will then absolutely look at that person’s account. We have people who are expert in self‑harm and suicide. If they are experienced, they can spot when somebody is joking. You do get people who joke about killing themselves because of the latest pop band splitting up.

Lord Sheikh: What about their football team?

Simon Milner: Or their football team losing. They are very good at being able to spot the signal from the noise. When they spot something that is really serious, it is not about getting that person to phone up the Samaritans; it is about getting the police involved. We have a network of relationships with law enforcement around the world, although not in every country. Certainly, we have very extensive ones with every single police force in the UK so that we can and do let them know. It is one of the few times when we will provide somebody’s data to the police or third party without their permission, saying, “We know about this young person. It appears they are at home. It looks like they have taken lots of pills”, and there are occasions when those people are rescued.

Lord Sheikh: It is a cry for help basically. They want to share it with somebody else.

Simon Milner: Yes.

Lord Gilbert of Panteg: That means that, when something is reported to you, a person has to sit down and review it. What is the triage system? How quickly in those most urgent cases can you deal with that issue, and how does it get to the top of the queue?

Simon Milner: It feels like that is a question for me, given we are hosting content. To give you a sense of the scale we are dealing with, we have about 1.8 billion people regularly using Facebook. Most of them use it every day. That is about 40 million people in the UK and, therefore, it includes lots of young people. That is a lot of people. We have billions of things happening on the site every day and millions of reports. We use technology to prioritise the most serious cases that may involve real-world harm, but we also ask good questions. We do not have just a big red button that says “Report”. If we do that, how do we know what the issue is? If you are not on Facebook, I would encourage you to join. Try using the reporting function. You will see that, depending on the nature of the content, we will ask you certain questions. The one that people typically might press is, “I do not think this should be on Facebook”. Then we will ask you why you think it should not be on Facebook. That will enable us to get that report to the right person quickly.

We have also found over years of experience that often a problem is not about our policies; it is about relationships. It is about a young person, or indeed someone of our age, who does not like a photo of them. It is not against our policy to post a bad photo of someone. We can provide tools particularly to give young people a form of words to ask their friend in an empathetic way, “I do not like that photo of me. Can you please take it down?” That means we do not look at it; it is not against our policies, but we have provided a way for those people to solve their problem, and then those kinds of reports do not clog up our systems.

Lord Gilbert of Panteg: How many times last year did you report to the police a very serious incident where you thought somebody might be suicidal?

Simon Milner: I am afraid that I do not have that number to hand. I am happy to check with my colleagues who work in that part of our business to see whether it is something we are able to disclose to the Committee. It will be no more than a handful.

Q111 Lord Sherbourne of Didsbury: Mr Milner, you talked about your policies and processes in place to protect the safety of young people and children. You say in your evidence that you are not disposed towards regulation because you find that your policies and processes are highly effective in safeguarding young users. What is the evidence for knowing that it is effective?

Simon Milner: It is principally from the reaction we get from our users. When people report to us, typically, we will ask them, “How was your experience?” in the way that many companies and organisations do, but it is also through our partnerships. I know that recently Tony Close and a colleague from Google, whose name, I think, was Ms Fussell, gave evidence to you. They talked about the different platforms and their experience of them. I think they were very positive about Facebook. I know Tony Close. I have not talked to him recently about Facebook, but to me that is a good vindication that the things we are doing are working.

We look to studies like the big EU safety work that Sonia Livingstone has led in understanding young people and their approach to safety. We look for independent sources of evidence as to how well we are doing, but, to be clear, we do not rest on our laurels; we do not think we have sorted out all the problems and can sit back. Far from it. We continue to invest, grow our expertise and look at whether we have got things right. For example, in the past year a number of our partners have said they think our safety centre is out of date; it is not accessible and does not work on a mobile phone. That may sound strange given how many things work on a mobile. Our safety team worked hard to revamp it, relaunch it and make it accessible, and change some of the language there and some of the advice to recognise changing technologies and the changing use of our service. We are always looking for ways to improve it, based on the feedback we get from both the people on Facebook but also the experts we work with outside.

Lord Sherbourne of Didsbury: If you do not get feedback where children have been badly affected by content, how do you know whether they have seen it?

Simon Milner: I do not quite understand the question.

Lord Sherbourne of Didsbury: Clearly, you are not complacent because certain things are not going right in particular cases. How do you know they are not going right? What evidence do you have for it? How would it be reported to you?

Simon Milner: Principally, it would be reported through our safety partners. We do not just ask our safety partners how we are doing; we also give them direct access to experts in our community operations team to let us know when they think we have made the wrong decision.

Lord Sherbourne of Didsbury: Are you talking about policies or outcomes?

Simon Milner: Actual outcomes.

Lord Sherbourne of Didsbury: Under your system, when children see content they should not be seeing, that is not always notified to you, is it?

Simon Milner: Let me give you a related example where something may be going on in a particular school. Often, waves of bad behaviour happen around an individual school community. It is something we have not seen before. People are reporting it, but we do not quite understand what the problem is because we do not understand the local context. That is when somebody like ChildNet or the NSPCC reach out to us and say, “This school has contacted us. A problem is happening in the school, but it is also on Facebook”, and they are not getting the right kind of outcomes they would expect from their reports. That can lead us to learn from that and realise—a ha!—that this is a new phenomenon we have not seen before. We now understand why we have been getting a number of reports like that. You are right that, if people do not report to us, we are not going to be able to take action. Therefore, it does need that young person or their friend to report it. You are right. Sometimes, we make a mistake; we have not reached the right judgment, and that is where the direct access that the safety partners have to our community operations team can help rectify the problem.

Q112 Lord Sherbourne of Didsbury: Can I pursue with Katie O’Donovan the Digital Economy Bill that is coming to our House very shortly? We had evidence last week from the BBFC. They told us that what is not in the Bill but might be is a clear legal requirement, which ISPs would like, for those providers to withdraw their services when publishers have not agreed to withdraw material. What is your view on that?

Katie O'Donovan: Obviously, the Bill is still in the House of Commons at the moment and has come under a lot of scrutiny. There has been a Committee in the Commons on it, and over the week the Government said they were now minded to request that ISPs block content. The Bill is a very good example of a straightforward ambition to address the issue of the availability of pornography, which has changed significantly since the advent of the internet. That is a sentiment we absolutely understand. One of our concerns is that the BBFC, the regulator, is very competent in understanding what pornographic material is and identifying that. You then have sites that are hosting pornographic material, which may be perfectly legal but are not meeting the requirement of age verification. The difficulty is that you have a regulator encouraging what are described in the Bill as ancillary service providers, perhaps payment providers or advertisers, to withdraw their services, but they are not legally compelled to do so. One of the things the Government have sought to do by including ISPs is to give a very clear legal direction to ISPs that it is for them to block access to that content.

Google is not involved in the value chain of pornography. We do not host pornography on our main platforms; we do not allow pornographers to advertise with Google products; and we do not host advertisements on pornographic websites. Therefore, it is not something in which we are directly involved in terms of the value chain, but it is a very good example of ensuring that, where there is legislation and regulation around the digital industries, it seeks to have a very clear purpose that is defined and discussed by the country, essentially. In this case, it is a parliamentary process. It is absolutely right that Parliament can ask ISPs to block content, but it is very helpful to have a clear distinction and direction of what content is legal and illegal, because it is very difficult for us in some instances as a content host or search engine to decide where a line should be drawn on content that remains legal.

Lord Sherbourne of Didsbury: But the BBFC could do that, could it not?

Katie O'Donovan: I think the Government are seeking to table an amendment, as they indicated at the weekend. At the moment, the BBFC can request ancillary services—those who support the functioning of the pornographic websites—to withdraw their services.

Lord Sherbourne of Didsbury: They can define what is legal, illegal, acceptable or unacceptable. They are well placed to do that, are they not?

Katie O'Donovan: They are very well placed to identify what is pornographic material, and that is very helpful. What is more difficult—I hope that the Government will bring clarity with their amendment—is whether content is illegal or legal. If it is adult pornographic material, which might not be suitable for a general audience but is still legal, it is very difficult to ask ancillary services voluntarily to do something. That is what the Government have sought to do where they have directed ISPs more firmly.

Q113 Lord Sheikh: Regarding the removal of content from Google and Facebook, it has been said that not all platforms are quick to take down offensive content. We have heard from children who have said that, in particular, Facebook does not respond adequately or quickly with regard to any complaints these people have made. How do you monitor and moderate content that may be unsuitable for children? That is my first question. My second question follows on from what I said a minute ago. How do you respond to complaints from children asking for content about themselves to be deleted? If you are going to do this effectively—in other words, to look at complaints concerning content—do you have enough resources? What are the benefits? What are the minuses regarding this approach in regard to responses?

Simon Milner: There are quite a number of elements to your question. I want to make it clear that we do not moderate content. That is principally for privacy reasons, but if you think about the scale of Facebook and the fact that most of what happens on Facebook is perfectly benign and in general is very positive, it would be completely inappropriate for us to be monitoring and moderating that.

Lord Sheikh: It may be benign to you but not to the person who has been offended.

Simon Milner: We do have very clear policies about what is and is not allowed on Facebook. One of them, which you raise, is that you cannot post an image of someone else without their permission. To that end, given that most people will not necessarily ask for permission, if somebody comes to us and says, “There is a photo of me that I do not want to be on Facebook”, we will take that off. As long as we are hearing directly from the person in the photo we will do that.

Lord Sheikh: Do you have adequate resources to do that?

Simon Milner: We certainly have an adequate process for doing it. For example, I complained this morning that my internet was completely cut off. I am hoping that by the time I get home it will be restored. If it is not, of course I will be upset about that. As with any organisation, you have to prioritise. We prioritise on the basis of real-world harm. We absolutely prioritise reports that come in from younger people on Facebook, but if it is an image they want to have removed that is not going to be as high a priority as, say, somebody reporting a suicide risk, or that they are concerned for their safety in some way. Therefore, there is a prioritisation based on that, but we try to get to these reports as quickly as possible. Every piece of content reported to us is looked at by a human being, and that takes some time. You have to ensure that the right expert who understands the language, our policies and so on is looking at that, but I want to be clear that not only young people but the parents of somebody under 13 can report an image and we will take it off Facebook.

We endeavour to try to get to things quickly and to be accurate in our decisions, but we are dealing with millions of reports every day. It is hard to satisfy everyone all of the time, but I do encourage you to try it. Try it and see what you feel about the response time, but also what you think about the message we give you. When you report something to Facebook, we let you know we have got your report. In some more complicated things we might tell you how long we think it will take, and then we will let you know our decision and why. We try to keep people, as per best practice across all service sectors, at least informed about how we are getting on with their report. I am not sure whether that adequately answers your question, but I hope I have addressed some of the points.

Lord Sheikh: Yes. I have one supplementary after Katie answers what I have put.

Katie O'Donovan: In a similar way to Facebook, if on YouTube there is a video that shows a child or somebody of any age and you do not want that video to be on there, you can request that it be taken down. You flag it and say, “This is my image and I am not happy with it. I do not give consent for it to be on YouTube”. We will remove it after we have reviewed it.

We also use the right to be forgotten for search listings. That was a ruling by the European Court of Justice in 2014 when it asked Google to de‑list and not return results for particular search queries of particular individuals who felt that the search queries being returned were outdated. If we agreed that they were not in the public interest, we removed them. Young people can use that as well. If those links to stories are no longer relevant and it is not in the public interest, we can remove those lists. Similarly, if there are websites that contain personally identifiable information, whether it is name and address or bank details, you can request that those results are not returned in search, which we will not do, and we do not do it for extortion sites either.

Lord Sheikh: If something is put on Facebook or Google that has a possible criminal element, because there are instances where these things do happen, are you proactive on that issue, or if something slanderous or libellous is said will you be involved in issues like that?

Simon Milner: With defamation-type issues, you absolutely have to rely on the individual who feels they are being defamed to get in touch with us. We have a special process for that, and we will act in accordance with legal advice in respect of any referral like that. There is one particular category of criminal content that we endeavour to ensure never reaches Facebook: child exploitation images. Those images tend to be illegal everywhere. Facebook, like most parts of the internet industry, uses technology to prevent known child sexual exploitation content to be shared via its servers. We are under an obligation as a US company to report instances of that to authorities in the US. We will not only close someone’s account; we will also report their details to the US authorities, which then share it with law enforcement elsewhere in the world to catch the perpetrators.

Q114 Baroness McIntosh of Hudnall: This is probably a question for Google entirely. When the European regulations that led to the right to be forgotten came in, what position did Google take corporately in advance of those regulations? Did you get behind them and push them along, or did you try to stop them?

Katie O'Donovan: We were very concerned when those regulations were first raised. The principal reason for our concern was that it gave us a responsibility that might be better served in a democratic or open process. The way that the legislation works is that, for example, if there is a news story about me and a petty offence I may have committed 10 years ago, I can make a request to Google for that search result not to be returned against my name. It is for our legal process and teams to decide whether that is in the public interest. That is quite a heavy responsibility for a private company to have, and it is one that we take very seriously. We were concerned about the precedent that set.

Since the legislation has come into place, we have been applying it as a European-wide piece of legislation, and we adhere to it. I think that in the UK there were about 100,000 requests last year. We granted 40% of those; in 60% of cases we said there was a public interest. This was primarily for adults rather than young people.

The systems we have in place are working well. We have a good system of legal review, and we have an annual report that makes clear all of the information we review in these circumstances, but it was definitely a concern to us when it was raised and, while it is working well at the moment, we keep it under ongoing review.

Baroness McIntosh of Hudnall: In relation to other requirements that might be placed on ISPs or other providers, do you still feel, in the light of your experience, that you would rather not have that kind of responsibility? Where would you look for those responsibilities to be taken up?

Katie O'Donovan: It has worked at a practical level, and we have additional policies. For example, we have a self-regulation policy on revenge porn. If you have been subjected to a former partner, or somebody else, sharing images of you on the internet and the internet host will not remove them, we de-list those and remove them from our search results. It is important that those conversations are had. They are real issues of great concern to people, particularly the more vulnerable in society. We now have a generation of people who have grown up online, which gives you a different set of challenges. It is absolutely right that policymakers across Europe and in the UK seek to find the best solutions to those issues. At the moment, the process is working, and we will continue to be part of that process, but we will also look for our own self-regulatory resolution as well—for example, policies on revenge porn.

Q115 Baroness Bonham-Carter of Yarnbury: My question for Kate is quite a niche one and is about journalism. I am interested in and have been concerned about content embedded in news items apparently from trusted outlets but somehow a bit of video has got in that is seriously unpleasant. What is your approach to managing that content? You might be a child who has Googled Iraq or something. It is a perfectly legitimate bit of wording and then there is something very nasty in the middle.

Katie O'Donovan: Is that on YouTube or on a third-party website?

Baroness Bonham-Carter of Yarnbury: It would be a Google search.

Katie O'Donovan: You are looking for a news story on Iraq and it is taking you to something else.

Baroness Bonham-Carter of Yarnbury: It is in the middle of the content, and I would be surprised if it had anything to do with whoever has posted it. Is that possible?

Katie O'Donovan: You can have scenarios where you are looking up Iraq, Syria or ISIS and you are looking for news-based results. You might go to a mainstream news organisation. You might go to slightly different editorially guided news organisations. Sometimes, within that footage, you can see content that can be absolutely distressing. It can have valid news value but would none the less not be suitable for a general audience. If you have opted for Google SafeSearch, which you can do from the home page very easily, those are de-prioritised in your search results, so it would be much more unlikely that you would have those results, but the content on third-party websites, whether it is a mainstream news organisation or a kind of citizen journalist website, is not something we have the ability to control. It is not within our gift to set the parameters of what that content should be. YouTube is our own platform and, therefore, we have strict terms and conditions about the content that is allowed. Graphic violence is not allowed on that. The exception is that, if there is news value to it and it is within context, we may allow that to remain, but often we will age-gate that content, so you have to be signed in and be over 18 to access it.

Baroness Bonham-Carter of Yarnbury: I see. You are using a kind of editorial function.

Katie O'Donovan: We do not editorialise search results. We give users what they are looking for. If you were to Google something like Iraq, it is more likely that you would get geographical information, information about population and that kind of thing. We will give the most relevant information to your search query, and that will be from a variety of websites.

Baroness Bonham-Carter of Yarnbury: Admittedly, I was not using safe mode, but I went straight to this page. It concerned me that young people, who might be given a project or whatever, could access it.

Katie O'Donovan: That goes back to the fundamentals that we are discussing here. Online, you have fantastic opportunities to read, experience and learn about different cultures and historical events, but there is also content that is absolutely not suitable for a general audience, so SafeSearch would be our technological response to that. We also feel that there has to be an educational and support response, so that when young people are using the internet they do it safely and understand that some things will not be what they expect to see. Some things will be unpleasant to see. They have to think about how they are searching and where they are looking for information and using trusted sources as well.

Baroness Bonham-Carter of Yarnbury: I obviously need that lesson too.

Katie O'Donovan: I think we all do at certain times.

Baroness Kidron: Do you operate the right to be forgotten in non-EU countries?

Katie O'Donovan: No.

Baroness Kidron: Therefore, there is a possible role for regulation. We had a brief conversation earlier about fake news on Facebook. What has come up a lot in our inquiry is the whole question of critical literacy, especially with regard to young people. We would love Simon to say something about that particular incident.

Simon Milner: What would you like me to say, Baroness Kidron? Do you want to ask me a question about it, or shall I just give you a general response?

Q116 Baroness Kidron: I think the question is: are you comfortable as an organisation with what has just happened with the fake news and the potential outcome of, and effect upon, the American election? Do you think it has further implications for other sorts of information that young people might be seeing and not judging clearly? Is that a sufficient question?

Simon Milner: That is perfect. Thank you very much. Our chief executive, Mark Zuckerberg, has spoken quite extensively about this issue since the events of 8 November, including at the weekend when he put a lengthy post on his Facebook page. I am happy to provide a link to the Committee for that so you can see the detail. I think there are three main points here. One is that our analysis shows that much less than 1% of the content on Facebook may be inauthentic, a hoax, fake or whatever words you use. We see no evidence to suggest that the sharing of fake news in relation to the US election made a significant difference to the outcome of it. Many other commentators have come out and said that as well since 9 November.

However, it is not a good user experience, particularly if people are sharing stories that are untrue. Therefore, we want to try to find ways to diminish the extent of that content on Facebook and reduce its prominence on people’s news feed, but we also do not want to be the arbiters of truth. There are many more experienced people in this room than me, but certainly during the years I worked at the BBC I rarely read a story about the BBC that I thought was wholly accurate when I knew the issue very well. That is not to do with the quality of journalism; it is just the nature of one’s understanding of an issue. You very rarely read something that is 100% accurate. Where do you draw the line if you are an organisation that is trying to depress the extent of that content, albeit it is already less than 1%, without making editorial judgments that people would think stray over the line, particularly for a platform that is used by 1.8 billion people globally?

It is an issue where we are determined to try to improve user experience, but also one where we are very mindful of the need to work very closely with publishers and those much more experienced than people in Facebook on these issues, and to learn from others about how best to tackle this. It is not an issue just about Facebook; it is a wider issue, and something where there needs to be a concerted and well-informed public dialogue about those matters.

Baroness Kidron: Does the 1% relate to the amount of news carried by Facebook or 1% of all the things posted on Facebook?

Simon Milner: That is a broad-brush number. It is less than 1%, which is anything from .99% to 0.001%. It is somewhere in that range, but we are not quite sure where—indeed, how would you measure it?—but it is about all content on Facebook.

Baroness McIntosh of Hudnall: Your point about things not being 100% accurate as reported is a perfectly legitimate one. We used to be told we should not believe everything we read in the newspapers. A lot of us were brought up in that way. Does it seem to you that part of your responsibility as providers as opportunities for these stories, whether they are wholly true, wholly untrue or partly true, is possibly to say to people, “You should not believe everything you read on Facebook”?

Simon Milner: I think this is one way where our community can be very powerful. Just because somebody has shared a story suggesting that the Pope was endorsing Donald Trump does not mean they were saying it was true. Quite often, people would share things and say, “Look at this utter nonsense”.

Baroness McIntosh of Hudnall: That is a very sophisticated nuance, if you will forgive my saying so, particularly for younger people.

Simon Milner: I am not sure I completely agree. There is a lot of evidence that young people are very good at identifying when something is wildly inaccurate. I agree with you when it comes to things that are plausible. It is plausible that Nigel Farage might be the next UK ambassador to the US. Who knows? When you have things that are not true but plausible, it is incredibly hard for anyone to judge. I agree there are some interesting questions about how you provide alternative perspectives, including fact checking. Katie may want to talk about the role of fact checking when it comes to news on Google, and we may need to learn from that.

Baroness McIntosh of Hudnall: I am sorry to press this, but I am asking you about something else. It is not about Facebook mediating whether things are or are not true; it is simply pointing out to people, as part of the normal way that you present what you do, that not everything that pops up and appears to have the authority of a particular brand, or even a particular design behind it, is necessarily information that should be trusted. It should at least be open to question. That is where the issue about critical thinking is surely most at risk, is it not?

Simon Milner: I absolutely understand what you are saying. One of the things that we do, for instance, is ensure that we verify certain pages on Facebook. If something purports to be from the BBC and it has a blue tick, it is from the BBC; it is not from somebody else pretending to be the BBC. That is one of the ways in which you can help people understand it. Generally, we look for brands that we trust. The verification tick that we, Twitter and other services use can be very helpful in guiding people about authoritative sources of information.

Katie O'Donovan: I think the critical thinking point is clear. In the past year we started a programme called Internet Legends, which goes into schools and works with eight to 11 year-olds. We were very pleased to have some Members of the Committee visit a school in Brixton last week. It is a general programme for eight to 11 year-olds, so those just beginning their online journeys. It sets out some very basic tenets and principles that we think it is good to encourage when people are online. One of the four is to check something is for real and ask whether, if you see something online, it is too good to be true. Is the source trusted? Think about it critically. That is a small way in which we are involved in starting that conversation. I know that there are many others active in the States. It is important for all of us—this goes for adults as well—that, if you can easily be duped by things online, you apply critical thinking, and we do have a role as a platform to support that.

Q117 Bishop of Chelmsford: Before I ask my question, I dive briefly into this debate. It is very important for the way young people, in particular, access the internet. Katie, a little while back in the conversation I thought I heard you say that Google does not edit material; it is just the pure noble provider of all of it. Am I not right in saying—correct me if I have got it wrong—that, when I put something into Google, the material it presents to me is not necessarily the same for somebody sitting next to me who puts the same thing into Google? What Google delivers to me is shaped very much by all the previous things I have been putting in as Google builds up its profile about the sorts of things I am interested in. First, is that true? Secondly, I do not think young people are sufficiently aware of that. That is editorial control. It is not the sort of editorial control newspapers provide, but it is editorial control, and we should be more open about that. I am not suggesting it is necessarily very sinister, but it is a fact, is it not?

Katie O'Donovan: That is a good and important question, and the answer to that probably has two very meaty sections. If you and your neighbour are not logged into Google accounts and using Google Search, you would get the same results. Those results would be based on a number of different indicators that our algorithms use to understand what would be the most relevant search. If you are looking for a news story, some of that would be indicators of the quality of the news site that we return. We use indicators like how many people link to this site. Is it linked to by other reputable sites? If you are a website that links you by the BBC, for example, that is understood by our algorithms to be probably quite a legitimate site.

Therefore, we have ways that are not editorial in the classic way newspapers behave in maintaining quality in the results we return, because that is absolutely what our consumers are using.

We do have ways of making sure that the search results that you look for are particularly relevant to you if you are signed into your Google account while you are doing the search. We do that because it is helpful, by and large, for consumers, but you are absolutely right that the trust and transparency that goes with that is key. For example, if you wanted to buy some wellington boots and searched for that, and you were signed into your Google account and that account was happy to have location features switched on, we might offer you wellington boot shops in London. If you wanted a take-away pizza, it might be in your local neighbourhood, but transparency is absolutely key in this.

We have a website called My Account and privacy settings within that, which we encourage people to review. You can personalise all the information we collect on you and how we use that information. Data enables us to give good personalised responses, which is usually what consumers want, but, if you would like advertisements that are not personal to you and not based in any way on your search history, you can turn that off. If you want to mute particular adverts, you can turn that off. It may be you were very interested in buying a car and have now bought one, so you want to mute adverts about buying a car. You can do that. It may be that you do not want us to keep a record of your search history at all because you do not want us to learn from that and improve and nuance your results. It is always confidential, but you do not want us to log that. Therefore, we have the ability to give personalised responses and usually that is helpful to the consumer, but we absolutely give the opportunity for consumers to turn off those elements.

Bishop of Chelmsford: I would love to pursue this further, but I do not suppose Lord Best thinks we have time for that. I just note that now we have become consumers, whereas the discussion was about how it affects other content that is not about being a consumer, but I will leave that one hanging.

Katie O'Donovan: “User” would have worked just as well.

Bishop of Chelmsford: Yes, but it is an interesting shift. What do you think is an appropriate age threshold for someone to use a social media platform?

Simon Milner: Shall I take this, given Facebook is a social media platform? We have a policy that for legal and operational reasons you have to be 13 to be on Facebook. That applies globally, with the exception of a couple of countries where it is 14 because of local law.

Bishop of Chelmsford: What do you do about eight year-olds on Facebook?

Simon Milner: There is nobody who will have their age set at eight on Facebook because they would not be allowed to have an account. I understand what you are saying.

Bishop of Chelmsford: We know that there is a barrier.

Simon Milner: The research done by Professor Livingstone and colleagues suggests there are not many eight year-olds on Facebook, but there may be quite a lot of 10 to 12 year-olds who are lying about their age to be on Facebook, which is of concern to us.

Bishop of Chelmsford: The Ofcom research suggests that there are quite a number of eight year-olds.

Simon Milner: The main things we do is ask people their age when they join Facebook. If somebody puts in their real age and finds they cannot open an account, we put a little cookie on their machine that means they cannot go back in and try again with a different age. That is one piece of technology that does it. We make this very clear in all the training in our safety centre and take every opportunity we can, including in front of this Committee, to remind people that you have to be 13 to be on Facebook, and indeed to use many online services. However, we have a fundamental issue in the UK—I know this was also evidence Tony Close and his colleague Lindsey Fussell presented to the Committee—whereby many parents choose, for whatever reason, to allow their children to go on Facebook. Often, they have helped their children get on to Facebook. When that happens, it is very hard for us to be able to know that that person is not the age they say they are.

However, we have special reporting processes. Whenever a teacher in a primary school says to me, “I have a real problem. All the 11 year-olds in my class are on Facebook”, I say, “You can report all of them to us”. Particularly when it is coming from a teacher I say, “Tell us about the accounts and we will act on all of them, because we know you are a trustworthy source”. They do not often do that because they think they might incur the wrath of the parents. It is a fundamental problem to which we have not found a ready-made solution. When millions of parents are making that decision, how can we enforce our policy? I do not like it; I do not condone it, but as a parent, all of whose children are now teenagers, I can understand why people might have made that decision. It makes it much harder.

Bishop of Chelmsford: Made what decision?

Simon Milner: The decision to help their child lie about their age.

Bishop of Chelmsford: We will not ask you whether you did with yours.

Simon Milner: I am very happy to say I never did.

Bishop of Chelmsford: Surprise, surprise: children lie about their age. I certainly did when I was younger. Whose responsibility is it to do something about this in your view? It sounds like there is a gap between your policy and what is happening. Should there be some regulation?

Simon Milner: There is regulation in the US and regulation is coming in Europe with general data protection regulation, which will be implemented by April 2018. Obviously, that will apply in the UK until we exit the European Union. Indeed, it may apply after that depending on the decisions the Government and Parliament take. It is not clear to me that that will fundamentally change the behaviour of parents, but it might. It creates another moment for the industry, together with government, safety organisations and so on, to remind people that these are the rules and why they are there. As per the guidance we have provided in our safety centre, that moment when your child becomes 13 is a great learning moment to say, “You can now come into a world that gives you incredible opportunities but also responsibilities”.

Bishop of Chelmsford: To turn the question on its head, why 13? I know that is not your decision, but, out of interest, why is that? You seem to be very committed to it.

Simon Milner: It is the law in the US. The Children’s Online Privacy Protection Rule has been there for quite a long time. I do not know the history of it and why the US authorities fixed upon 13 as the age. I expect Professor Livingstone knows better than I why they went for that particular age, but I am happy to look into it and provide anything further that I can.

Baroness Kidron: Simon, has Facebook put any money, thought or creativity into some sort of public service campaign—I cannot think of a good way of saying it—to parents saying, “Do not do this”? Is that something you could or might consider?

Simon Milner: We certainly have supported safety organisations who make it an important part of their work to ensure parents know this rule.

Baroness Kidron: But not within the Facebook ecosystem itself.

Simon Milner: If you think about it, once people have made that decision they are allowing their children to be there. It is not obvious that it is going to make much difference. We feel it is a situation when perhaps parents are all together getting that talk on internet safety. For those of you who have had children, nine, 10 or 11 is often the time when schools are doing quite regular internet safety training with the likes of ChildNet, NSPCC and so on. When all parents are in the room together, that is a great moment; that is a good learning opportunity. We provide both financial and other support to those organisations when they are making those kinds of interventions.

Lord Sheikh: May I put a very brief supplementary?

The Chairman: Lord Sheikh, we have finished our hour of conversation and are just coming up to the halfway mark, so I am not going to have any more supplementary questions, sadly. We have to move at a bit of a pace. Could colleagues ask their questions briefly, and could our two witnesses, very kindly, give pretty clipped, tight responses?

Q118 Baroness McIntosh of Hudnall: These are questions about privacy settings and your attitude to how rigorous you should be about this. The UKCISS guide states that the company should offer privacy setting options, including privacy by default, to give control to its users. Do you comply with this? Do you offer default-on privacy settings and, if not, why not? I have already asked you about the right to be forgotten. This is a matter more for Google than Facebook, but, in the event that those regulations do not apply because we are no longer part of the European Union, would you cease to apply them?

Katie O'Donovan: I can start with that question and work backwards. We have not made a decision about that yet.

Baroness McIntosh of Hudnall: Therefore, it is not necessarily “yes”.

Katie O'Donovan: It is an honest answer: we have not considered that and, like everybody else, we are looking to see what happens as the UK prepares to leave the European Union. Obviously, the Government have said they will transpose GDPR and lots of other legislation. From what we have gathered so far, it looks like the right to be forgotten would still apply, and we would happily concur with that. If the UK Parliament decides something else, we will engage with and work with it to ensure that the alternative is helpful and productive and does what Parliament wants it to do.

As to our privacy settings—I mentioned some of this in my previous answer—we believe they work for our users to enable them to have a personalised and effective use of our services. That means some of them are on by default, and we give people access, and the opportunity, to change those very easily. We make sure that our terms and conditions are easily understandable. We work as a cross-functional team to make sure they are not legalese jargon and not too long. We were very pleased to win an award from Time magazine for clear terms and conditions.

We also make sure that we respect people’s data related to different age groups. We have different priorities for people aged 13 to 18 who use our services and a different approach to advertisement in the way their data is held. We make sure that all data, whatever the age of the user, is never sold or passed to third parties; it always stays within Google.

Simon Milner: We comply with the UKCISS guidelines. As I mentioned earlier, whenever anyone joins Facebook their privacy settings are set to the tightest level, which is friends only. That will not change unless and until they change them, and that is true for all age groups.

Q119 Lord Gilbert of Panteg: Can we move on to the issue of privacy of data that you collect from your users, particularly children? This came to public attention when Facebook and WhatsApp were proposing to share data, but it is a wider issue. Can children, or any user, decline permission for you to share data with third parties and still use your services?

Simon Milner: Yes. We do not share people’s data with third parties without their permission. That is true of everyone, however old they are, with the one exception of law enforcement requests. When we get such requests we can and do share data. We make public the number of times we do this. We report this every six months. That is the one time when we will share people’s data without their permission, if there is proper legal process and evidence of a crime and so on.

Lord Gilbert of Panteg: Do third parties include other companies within your group?

Simon Milner: No. They are part of the Facebook group. Depending on the actual data policies of those companies, we can and do share data between the companies within the group.

Lord Gilbert of Panteg: Could a child, or somebody else, decline permission for you to do that and still use your services?

Simon Milner: It depends on which particular services we are talking about in terms of those within the group. Rather than go into it now, I am happy to write to the Committee afterwards to explain the details of that.

Lord Gilbert of Panteg: Is it the same for Google?

Katie O'Donovan: We do not share the data of anyone of any age with third parties. We do not share personal identifiable information with third parties. You can choose to decline to share any of your data or usage with Google and still use all our services in exactly the same way as if you were fully sharing it.

Lord Gilbert of Panteg: Do you think children understand the extent to which you analyse data about them in creating your profile of the user?

Katie O'Donovan: We are very keen to ensure that more and more people do understand the way that we use data to make sure our services deliver effectively for them. We launched Privacy Checkup on My Account. We regularly advertise that to our users and encourage them to have a look at their activity. I think we have had about 1 billion visits since we launched it. Every year, on Safer Internet Day, we have home page promotions on Google to direct people to that section of the website. We have offered free data to people to have those check-ups. It is a very easy to understand and intuitive guide to what you want to share with Google and how, and that is available to anyone who is 13-plus and has a Google account.

Lord Gilbert of Panteg: Do you think children understand how Facebook uses their data?

Simon Milner: We are endeavouring to help them to do that. I rather look to third parties to give you hard-and-fast data on it, but we try to make our data policy more accessible to everyone. Last year, we relaunched it. It is about a third of its previous length. We changed some of the language to make it more accessible and had clearer sections in it. We do not have a separate version for children. However, we do work with organisations like ChildNet, a long-term partner of ours. They have produced their own guidance on how services like Facebook and how advertising work on Facebook and so on. That is probably the best way to try to help young people understand how services like ours work.

Q120 Baroness Kidron: One of the things we have been trying to look at is the idea of well-being rather than safety and expanding this conversation, because safety is not enough. Simon, I think you have already referred to it. One thing we have taken a lot of evidence on is that the always-on culture is a problem for kids, particularly at the developmental stage when they are not entirely in control of their own feelings. Maybe you could give us a response to some of the evidence you must have seen about screen time and the compulsion to respond and share.

Simon Milner: It is certainly something we are aware of. We talk about it a lot with our safety partners. One thing we are very clear about is that what is right for each family can be quite different in terms of screen time. What is right for each member of the family when growing up can be quite different. What worked for my 19 year-old daughter when she was 14 does not work for my currently 14 year-old boy.

However, we also say it is important for parents to set a good example. If they say there should be no phones at the dinner table that includes parents. That is part of the guidance. Often, young people will copy what they see from their parents in terms of use of technology, so setting a good example can be very important, but try not to preach. Who are we to tell parents how to bring up their children?

We would also say that when it comes to well-being—I am very pleased to hear this is an important part of what the Committee is looking at—we know from experts that social media can become incredibly important to people when they are under stress. When they are mentally challenged, it helps to bring people together who are suffering from the same condition, if you like. Not all young people have happy home lives. Unfortunately, we hear all the time about young people who are in the most stressful situations at home, not when at school. Therefore, having the ability to access secret groups on Facebook, where their parents do not see what they are doing, can be very important for those young people. Like all these things, it is not linear and it is not a one-size-fits-all approach. It is about ensuring that we try to provide the best general advice we can to the great majority of young people and parents who live the same chaotic lives most of the rest of us do but generally are getting on with it and finding a way through. We also work with specialist experts, whether it is CEOP, the NSCPP, Lucy Faithfull Foundation and others, who deal with the most vulnerable children and those under most distress, for whom social media can be a lifeline. Getting that balance right can be incredibly hard, but we try.

Baroness Kidron: I acknowledge that what you have just said is absolutely true. It provides a huge resource for young people, but most platforms, however, are designed to extend use. One thing that has come up again and again is young people’s feeling of being overwhelmed by alerts and notifications. Should some of these things be off by default? Should some of them give them a break? Are you comfortable with the design of your services with regard to the developmental stage of the people who are using them?

Simon Milner: I can tell that Katie wants to come in. I am happy to come in if there is time.

Katie O'Donovan: I am also conscious of time. We would echo a lot of the points Simon made. The LSE and Sonia did some very interesting research recently on the differentiation of different types of screen time and how we cannot lump it all together as bad, which is the point you are making.

Your point about technology and whether we can build some of that into products is something we are very interested in as well. When we launched YouTube for Kids, we developed a time limit so parents can choose how long they would like their children to use that for. As users of technology get more and more sophisticated, obviously they have the ability to override those as well, but the elements we can build in by design are really important. That is something we think carefully about as a company.

Q121 Earl of Caithness: Notwithstanding everything you have said, the problem for young children is getting worse because of offline and online access. There is a real problem here. You can help to control that by the design of your platforms, so could both of you give an example of what you can do to your platforms to make it safer for children?

Simon Milner: That is a very general question. What have we learned to do? We have learned to put help where it is most needed. We have a very extensive help centre and very much encourage people to go there with their problems, but the key thing we have learned is that nothing is as good as inline help. When you are experiencing Facebook on your mobile device or computer and see something that distresses you, there is help right there when you need it, or, when we start to see behaviour like public sharing, which is an unusual pattern for you, we intervene at that time to say, “Are you sure you want to be posting publicly?” I think that the more we can provide inline, real-time help, that is very effective, in the same way that, when you talk to teachers and parents, it is all about, “I really wish I had been there at that moment, because that was the teaching moment”. We are always looking for teaching moments as part of young people’s experience on our platform.

Katie O'Donovan: The answer I would give is that most specifically for younger people it is, first, SafeSearch and restricted mode, which is restricted search on YouTube as well. That enables people to experience the full possibilities and opportunities of the internet but in a more restricted and safer way. Secondly, with YouTube for Kids, we took the decision to invest a lot of time and engineering effort—it was much more complicated than we thought it would be when we set out—into building a product specifically for younger people that does not collect their data or personal information and does not allow them to communicate. It has a very in‑depth onboard inflow for parents that enables them to learn about how to report and flag content on YouTube, which is really important.

Earl of Caithness: Let me turn it around to you, Katie, and then Simon. Why do you not have on Google an automatic stop after you have done a YouTube search instead of continual running of the next YouTube video that you do not want? Think about that one. Simon, why do you not have a device, or alarm system, which says, “You have been on Facebook for an hour. Is this good for your health?” Those are two things you could do easily. Why do you not do them?

Katie O'Donovan: On YouTube, we are trying to serve content that is relevant to the viewer. If they have watched a video and we offer one that is similar and they do not want to watch it, they can very easily click pause or come out of the app and stop watching.

Simon Milner: It is extremely unusual for anybody to spend that length of time on Facebook. The average amount of time over a whole day might be an hour, and that will always be in fits and starts. It is an interesting idea. I will happily pass it back to our safety team to see what they think.

Earl of Caithness: Cut it down to half an hour.

The Chairman: There are two good ideas. The fact we have overrun by miles is a tribute to how much we have learned from you. Thank you very much indeed for all of that. I am afraid we did not get round to asking you about advertising on which we would be very interested to hear you. I am afraid we must do that by correspondence. Email would be just right here. Thank you both very much indeed for joining us.

Examination of Witnesses

Dr Sarah Marsden and Dr Akil Awan.

Q122 The Chairman: Dr Akil Awan and Dr Sarah Marsden, you are both very welcome. We are sorry to have held you back for 20 minutes, but we were having a very important discussion. If I may, I will ask you to introduce yourselves as we move on to the issue of radicalisation through the internet. Perhaps you would kindly tell us a little about yourselves and make any opening remarks that you wish.

Dr Sarah Marsden: Thank you very much for the invitation to be here. My name is Sarah Marsden. I am a lecturer in radicalisation and protest in a digital age at Lancaster University. Prior to that, I was a lecturer in terrorism studies at the University of St Andrews. I have spent the last 10 years or so researching terrorism and political violence, with a particular focus on engagement with and disengagement from violent extremism and militant Islamism.

Dr Akil Awan: Thank you for the invitation. I am Dr Akil Awan. I am a senior lecturer in modern history, political violence and terrorism at Royal Holloway, University of London. Over the last 10 years I have been looking at radicalisation, social movements, the role of religion and the history of terrorism, broadly speaking. My most recent work involves looking at terrorist propaganda. I also work with the United Nations on youth and radicalism, particularly Security Council Resolution 2250, which is all about youth.

Baroness Quin: This is very much a get-the-ball-rolling question. Thinking about the triggers for radicalisation, are young people more susceptible to being influenced online? If so, why do you think that is?

Dr Akil Awan: The brief answer is yes. To expand that, the internet, in particular social media and web 2.0 platforms, has emerged as the principal arena for youth engagement, politically and socially, over the last decade or so. That is largely a positive thing. It is conducive to egalitarianism and levelling of the field, if you like. There is a slight problem, in the sense that principally it is a function of young people being what are thought of as digital natives, as opposed to digital immigrants. Digital natives are those who are born into the digital world of computers, the internet, video games and that sort of thing. Conversely, the rest of us—I am sorry to point out how incredibly old all of us are—are digital immigrants; we have had to come into that world, sometimes kicking and screaming. We appropriate the language of the digital world, but not in the same way.

For that cohort of young people, any real-world activity, whether it is shopping, playing games, dating, reading or socialising, has a virtual counterpart that might be more appealing. Therefore, it should not be surprising if their political activism or radical escapism also takes place within that sort of arena, so in part it is because of their immersion in that sort of environment.

Baroness Quin: Sarah, do you wish to add anything?

Dr Sarah Marsden: It is important that in my comments I enter the caveat that I am not an expert in child psychology. My comments refer to the literature on radicalisation in respect of the question. The years up to 18 include a lot of variation. There will be a lot of variation in susceptibility and vulnerability along that age continuum.

In general, younger people have less digital literacy and fewer critical consumption skills. They have not necessarily had the opportunity to develop those skills, so the messages they receive may be critically engaged with to a lesser extent than for those who have received training or have had negative experiences. In addition, that period of adolescence, when people seek an understanding of themselves, how they interact with the world and their place in it, brings with it a series of vulnerabilities that are important to bear in mind. I reiterate the point that engagement with the online space can be a really positive thing for political activism and for learning about the world and about political, civic and social questions.

I reiterate the point that young people engage in and experience a lot of things that we might consider risky, but they do not necessarily suffer as a result. There is resilience in most individuals, so it is a case of balancing potential susceptibilities with recognition of the resilience that already exists and the extent to which different people will be affected in different ways.

Q123 Lord Sheikh: This subject is of great interest to me. I prepared a report on problems relating to Muslims, which has been sent to the Prime Minister. I also spoke about it in the House of Lords. We are talking about radicalisation online. There are a number of ways in which people are being radicalised, but we will confine ourselves to how things are done online. There are about 3 million Muslims in this country, nearly all of whom are doing well. They are peace-loving people who have contributed to the advancement and well-being of this country. A lot of them are entrepreneurs like me, some of them are here, some are doctors, and so on, but we have a problem with online radicalisation. I have gone up and down the country and talked to imams. I have addressed meetings in Birmingham and Manchester. Online radicalisation is an issue that crops up.

We are limited by time, so I am going to come to the crux of my questions. First, has the internet allowed for an increase in communication by radical groups? Has that expanded because of what is available on the internet? How do those groups target young people in particular? How do they recruit them? It is not just propaganda from this country but what comes from other countries as well; for example, what happens in Belgium or France. As you very well know, this seems to be an international problem. Those are the points on which I would like information.

Dr Akil Awan: On the increase in communication online, I should point out that the idea of online radicalisation as the most important issue facing us has largely been debunked in academia. There are no real cases of completely autonomous online radicalisation. Children live in online and offline real worlds simultaneously, so in a sense it is pathologising the internet a little bit.

Having said that, one of the paradoxes of the new media environment, broadly speaking, is that we have before us a wealth of information in the form of various languages, various viewpoints and various ideological perspectives. Increasingly, we tend to silo ourselves into the particular viewpoints that seem to corroborate or confirm our world view. We are all guilty of that in our own news media practices. I use the BBC as my touchstone, generally speaking. Therefore, there are spaces online where young people in particular find themselves cocooned, and they end up in a kind of echo chamber. The same sorts of views are reiterated ad nauseam, in effect, without debate, dialogue or challenge. That environment is very conducive if you are trying to bring someone round to a particular way of thinking.

We could add to that the issue of the long tail of the internet. That term is taken from marketing or business, but the idea is that if you are a young person in a community and you hold particular views, or are inclined towards particular views, you will probably not meet someone who shares those views in your school or community. However, online you are far more likely to find someone who corroborates those views and, by some sort of reciprocal legitimation, you end up reinforcing each other’s beliefs. That is an issue. Of course, the internet allows you to do that anonymously. If you were to express those views in a community, you would open yourself up to scrutiny from law enforcement, parents and teachers, but also perhaps to ridicule and satire. Online you can be anonymous. We have seen cases of people pretending to be something they are not as well.

Dr Sarah Marsden: I would like to reiterate the first point you made. The scale of the problem is small, in the sense that relatively few people actually move towards violent extremism. Since 2001, maybe 500 people have been convicted of terrorism offences in the UK. For the under-18s, it is a very small number, about 3%. You are dealing with hundreds, not thousands. That is an important point. When you think about the implications of responding to radicalisation and the online environment, it is very important that proportionality of response is kept front and centre.

In answer to the second point, as for everybody the internet has extended the boundaries by which we can communicate. It increases the number of people we can talk to, the types of people we can talk to and the sorts of views we encounter. Of course, radical groups use that facility for propaganda to promote their particular ideas; they use it to communicate with and to recruit, in an effort to try to bolster their cause, people who are already immersed in extremism or are in a particular radical setting.

Lord Sheikh: How do they establish whom they should target?

Dr Sarah Marsden: If I may just finish the first point, it is important to bear in mind that there is a range of factors in the way radical groups communicate. It is important to bear in mind, too, that the internet is not without constraints for radical groups. There are constraints that operate on them. The capacity for trust in online settings is constrained because you do not know who you are engaging with, and there is a high degree of suspicion that most people are from the CIA or MI5. Alongside that, online venting and communicating in extreme ways might not necessarily move somebody towards violence. It is probably not reasonable to assume that there is a process at work all the time. Availability of messages does not necessarily equate to impact. Just because something is available does not mean it is necessarily having an effect on people. We come across all sorts of media advertising that we do not necessarily engage with.

To respond to the second point, research on this particular issue, which it is fair to say is limited, has identified a range of ways in which people will try to seek out individuals who might be of interest to them and who might be interested in radical ideas. For example, on social media they might hijack particularly popular hashtags and include their own content as a bridge to start communication. They link to other radical accounts that are just this side of legal in order to try to identify a wide number of people who might be interested in their ideas. Where people respond—most obviously do not—there is evidence of an attempt to create what has been described as a micro-community. A huge amount of effort is put into trying to engage with individuals who show some kind of interest. Then there is a move to the offline space, or at least the encrypted space, so that communications can take place out of the public eye. Beyond that, there is an effort to try to move people to action.

Lord Sheikh: How do we combat what is going on? What is your remedy?

Dr Sarah Marsden: I do not have a remedy.

Lord Sheikh: You know we have a problem.

Dr Sarah Marsden: Yes.

Lord Sheikh: For instance, what do you want people like me to do?

Dr Sarah Marsden: First, keep that question of proportionality in mind. Positive responses are far more appropriate than negative ones, in the sense that, rather than taking down content all the time, we should try to develop individuals’ digital literacy and critical consumption skills. Find alternative positive ways of engaging in civic and political activism. In my own work, in the effort to support people moving away from extremism who have already been involved in radical settings, it is far more effective to find positive ways of directing that initial motivation to become involved in extremism in the first place. If the motivation is related to group belonging and relatedness, finding alternative ways of supporting that is a really helpful way forward. Similarly, if people are genuinely committed to questions of social justice, finding positive ways of pursuing that seems more appropriate to me than a cat-and-mouse game of taking down content and censoring what is available to people.

Q124 Lord Allen of Kensington: Turning to logistics, is the internet being used as an online recruiting opportunity? We have seen press coverage that Daesh published a guide about how to get to Syria, safe houses, flights or whatever. Has the increased use of the internet enabled it to be much easier for people to access logistics information and find data that take young people to conflict areas? If that is the case, is there something practical that we can do to deter them?

Dr Akil Awan: That has certainly been our experience. Sites such as ASKfm, JustPaste.it and many other social media sites are almost catering to the desire to know about how to go about engaging in that sort of thing. If you are searching for flights online and looking at how someone else managed to cross the Turkish border, for example, you have a precedent. Some of that also happens in the real world, in peer networks. Some of the individuals who have gone to Syria come from the very same town, despite not being numerically or in any other way significant.

There is one other thing the internet has allowed, over the last couple of decades at least. Typically, a crime needs both motive and means. The means have become a little easier, in the sense that there is a whole swathe of DIY manuals and all sorts of instructables on the internet that allow individuals to go beyond the idea of visiting a training camp or having real-world training. Often, they are not very good. That is one of the reasons why we have not seen many success stories for wannabe terrorists who follow DIY guides, for example. There are pros and cons as regards the leaderless or lone-wolf-type attacks.

Dr Sarah Marsden: I agree that logistics have become facilitated by the internet. There are barriers to mobilisation, particularly for young people, such as parental oversight, passports and financing it. There are constraints on radical groups about the extent to which they can facilitate it in a practical way. Although there are reports of people receiving money and tickets, there are still limits on the extent to which that is feasible. I agree that the limits of the internet as an instructional tool are significant. The internet is not necessarily a virtual training camp; it is hard to create viable bombs, and we have seen people fail as a consequence of trying to do so via the internet. Those real-world experiences are vital to the move to violence.

Earl of Caithness: Dr Marsden, to follow up an earlier reply you gave, do you have any evidence that the attempt at radicalisation is increasing or decreasing on the internet?

Dr Sarah Marsden: The short answer is no, I do not have an indication of whether things are increasing or not. There is some evidence that the internet has not necessarily increased violent extremism. The most recent research on that has tracked the relationship between the number of people in the UK or the West who have become involved in violent extremism and the development of the internet. There has not necessarily been a correlation. It is not a case of saying that the internet has caused an increase in terrorism per se. That is probably as far as I would be confident in going.

Q125 Baroness Benjamin: You said earlier that the numbers who become radicalised are quite small, but those who do can destroy the lives of many. I am sure you agree that we all need to be vigilant, and to be aware of those who have become radicalised. How engaged is industry, such as the ISPs, search engines and social media platforms, in tackling online extremism?

Dr Akil Awan: There is a range of measures. Of course, we have the take-down policies and the filtering, some of which we heard about today. Some of the social media platforms, such as Twitter, have been very active in taking down hundreds of thousands of accounts. There is always a tension in whether you go in with a heavy hand and take down everything. There is some value in having the presence of those sorts of organisations in public fora in a sense, because they provide intelligence; it also prevents them going underground, in effect.

Another point I want to bring up relates to the recent election in the United States. It is not really about online extremism per se but it is related to that: the circulation of news and stories that are not necessarily extremist per se but are conducive to creating an enabling environment that results in the growth of those sorts of problems. I’m not quite convinced I heard a reassuring answer earlier from the Facebook representative about fact checking. A number of reports have spoken about the effect some of these stories have had on the post‑factual era we are now living in. “Post-truth” was voted word of the year by the Oxford English Dictionary. We have to be aware of the challenge this now poses for us. If you take a tabloid newspaper, there is still some recourse to the PCC, or IPSO now, but in this case, when a large proportion of individuals are getting their news and form their world view based on completely fabricated stories, what responsibility do social media platforms, which allow those stories to circulate, have in response to them?

Baroness Benjamin: Do you think there is a legal case for taking them to court?

Dr Akil Awan: It is outside my purview, but I do not think so. We need to be a bit more creative about some of the solutions we propose. This is a relatively new problem. We were not talking about it a year ago.

Dr Sarah Marsden: Industry is increasingly engaged. As Dr Awan says, information is taken down; millions of videos have been taken off YouTube; hundreds of thousands of Twitter accounts have been closed down, and so on. They have trusted flaggers who identify inappropriate content, but innovation and thinking creatively is key. Google is doing a little bit in that regard to bolster positive voices. The online civic courage initiative, which has been developed with some think tanks in the UK, has tried to facilitate and support counter-speech—I am not sure whether that was covered earlier because I was not here; it is emboldening people to engage, criticise and debate the ideas people present online.

The effort to try to support that process so that information and ideas are challenged constructively is useful. However, it comes with the caveat that it needs to be done carefully and thoughtfully; it needs to be transparent, because the wider question about industry’s engagement with it is the extent to which we wish our online experiences to be manipulated for what are political ends. Who is most appropriate to do that is a really important question. It demands transparency and thought about how we engage with that question appropriately. It is not the same as child sexual exploitation where you can easily take down content. It is a political question. There is a spectrum of political content online, so trying to think about what should be removed, or how it should be engaged with, is a challenge, and transparency sits at the heart of that.

Baroness Benjamin: Do you think filters are effective in stopping extremism? Are they working? How do ISPs effectively detect extremism in a foreign language and in foreign form?

Dr Sarah Marsden: The short answer is that you would probably be wiser to ask Facebook and Google. I do not know the answer to the particular question about language. My understanding is that most of the social media companies rely on community reporting, primarily working on the assumption that people in the community identify content that they think is inappropriate. That is flagged and then reviewed by a human being and taken down. There are trusted flaggers who are part of that process, some of whom are state actors such as the police and so on. I do not know how effective it is. The research on it is nascent. I would be cautious about saying how effective it is, because there are unintended consequences. The move to encrypted spaces, or spaces that are less protected online, is a potential negative consequence. It is also important to bear in mind intelligence gathering and surveillance. There are pros and cons, and, as that question is explored further, it is important to keep a balanced view of both the potential positives and negatives.

Baroness Benjamin: What more could the industry do to monitor and prevent access to extremist material?

Dr Sarah Marsden: It is very difficult. My response tends to sit around developing young people’s critical thinking and political consumption skills and their digital literacy so that they are better able to assess and interpret the content they find online. Taking things down and responses that sit more within the censorship space are less desirable than developing resilience in young people.

Dr Akil Awan: I corroborate that. There is a whole host of things we might do that are not necessarily about content but about how young people engage with it. That is the key. We have to think about the role schools play in developing critical thinking and young people’s abilities, and teaching them methods and skills so that they can weigh evidence and contextualise the knowledge they receive, and engage with it in a much more sophisticated and nuanced way. That is fundamental. It involves digital literacy and media literacy, too.

There is a whole host of other things that we could do. Perhaps we are entering the schools section. One of the things we are working on at Royal Holloway at the moment is a project about teaching the crusades in schools. The crusades are used quite problematically by a whole swathe of extremists from all sorts of different political and ideological spectrums. The bin Ladens of the world talk about themselves as chivalrous knights fighting against the crusaders. Anders Breivik, the Norwegian terrorist who killed 77 people in 2011, talked about himself as a Knight Templar, a modern-day crusader resisting the incursion of the new Muslim waves. Both of those are reciprocally legitimating in a sense; they work towards the same outcome, which is a very unhelpful thesis on the clash of civilisations. They are in a sense more than the sum of their parts. Young people, as far as I understand it, do not encounter the Crusades. I certainly did not study the Crusades at school. I think that at key stage 3 you have the option of coming across the topic. If you encounter very tendentious, skewed and polarised views of the world, for example about history, you do not have anything to push back on; you have no resource to challenge and contest that in any way. If we were able to teach some of these topics in a much more nuanced way, making them compulsory in part of the curriculum, we would provide students with the tools to be able to push back on some of those things. We would almost be providing some kind of inoculation against very poisonous messages.

Another crucial thing that we might do in schools is promote progressive and inclusive ideas of citizenship, belonging and civic identity. We tend to think that political socialisation begins when children turn 18, but it does not; it is a continuous process. If someone feels that the Iraq war was unjust and carried out for ulterior motives, you might not be able to change their mind cognitively but you might teach them how they might respond. Teach them that there are other ways. Teach them political socialisation and how they might be part of their community and the political process. That can be quite helpful. If you teach those sorts of things at school, you are dealing with some of the problems quite early on.

Q126 Bishop of Chelmsford: You anticipated some of the questions I was going to ask about schools, but what you say is very interesting. It is interesting to me that at a time when religion, often for all the wrong reasons, is driving this conversation and is on the front page of newspapers, in other bits of the world religious literacy, and the more nuanced understanding of the things you are speaking about, is taking such a back seat. Religious education is not part of the national curriculum. We have just heard in the past week that the BBC is further cutting back its profile of religious broadcasting. Aaqil Ahmed has lost his job, and there is now nobody in the BBC with any real knowledge of religion working as a commissioning editor. That question is rather wider than this topic, but it is very relevant to the place of schools in education to help people get a world view that might give them other narratives with which to interpret and deal with some of the stuff they encounter.

Sarah, you may want to say more about the role of schools, and this question is very specific: given that the Government’s Prevent strategy places a high responsibility on schools to do something, do you wonder whether schools have ended up being overcautious? Has it gone the other way? I would love to hear your reflections on all of that.

Dr Sarah Marsden: The Prevent duty has certainly presented challenges to schools. One of the primary ones is securing schools’ place primarily as a safe space for educating young people. The challenge there is resisting the increasing security agenda being presented to schools as part of their responsibilities. That is not to say that safeguarding does not matter; of course it does. It is not to say that safeguarding is not an important part of a school’s role, and extremism is part of that. Front and centre, schools are places for education, the development of thought and young people’s well-being and flourishing, and sometimes the security agenda can get in the way of that.

I agree that to some extent, with the Prevent agenda and the demand that schools report people at risk of radicalisation, there is a tendency to be overcautious. We are only at a very early stage of the policy rollout, so the full effects will play out over time, but there is a risk that people will be overcautious and assume that things are problematic when they are not. It takes courage for educators to be conscious of their responsibilities to put education at the heart of all those questions. If somebody is asking difficult questions, that is fundamentally an educational issue, not a security issue first and foremost.

The capacity to resist that demands also that teachers and schools have the training, resources and skills to feel confident about having what are potentially very difficult conversations. That is why it is important that schools remain safe spaces for those difficult conversations. The implication of that is that young people not only develop knowledge about particular historical events—those critical thinking skills are very important—but think about how to engage with difference in an increasingly globalised world, so that we learn and teach how to engage in a robust pluralism. We will not all agree, so how do you engage with young people and enable them to have those difficult conversations with their peers and parents when they do not necessarily agree? It is also about enabling and providing mechanisms and support for young people to explore those questions and resisting the securitisation of the educational space.

Dr Akil Awan: The Counter-Terrorism and Security Act 2015, which added to the Prevent duty, creates a statutory duty for those in the public sector to prevent radicalisation. Essentially, it puts the onus on them to identify the early signs of radicalisation. It has been quite problematic for all sorts of reasons, particularly for children. There has not been sufficient training of those in the public sector about how they might spot the signs of radicalisation, whatever that might mean. Just before I came, I printed off the Government’s vulnerability assessment framework, which is one of the things they use to identify the signs of radicalisation; I want to refer to four that relate to young people in particular. One is being at a transitional time of life; another is a need for identity, meaning and belonging; another is a desire for excitement and adventure; and another is a desire for political or moral change. There are others, but those things could apply not only to every student in this country but to every child on the planet. When there is such a broad swathe of risk factors, or things that might predispose you to recognise a radical, it becomes almost nonsensical.

We have seen a number of referrals under Prevent, and they do not get followed through, because we are talking about false positives. And really what we are doing then, is looking at markers of religiosity, and often ethnic-based, which is one of the reasons why the National Union of Students has refused to implement Prevent. That is the kind of problem that results from that intense focus. It is not very helpful.

As Sarah said, young people need to be allowed to explore ideas, particularly at university. There is now a statutory duty on university lecturers like Sarah and me to report anything that could tick some of those boxes, but universities are supposed to be safe spaces in which you can explore ideas, and make mistakes politically and ideologically. There is the famous adage that if you are not a radical when you are young you have no soul, and if you are still a radical when you are old you have no sense. That is part of our political socialisation. We go through that process as young people.

Q127 Baroness Kidron: I was going to raise the issue of Prevent, too. There is one particular criticism I would like both of you to comment on very briefly. I refer to the anxiety that parents are unable to discuss issues at home; it creates the fear that kids will then talk about them in educational institutions and be noticed for that. Have either of you come across that phenomenon? It seems very interesting to me. It came from the UN special rapporteur. The other thing is whether you have any suggestions for government, given that Prevent is somewhat problematic. What would you like to see from government particularly as it relates to young people?

Dr Sarah Marsden: I have seen evidence of the alternative conversation, with parents saying to young people, “Do not discuss these things because you could into trouble”. From a Muslim parent’s perspective, somebody described it as, “I’m going to have to have that difficult conversation with my 12 or 13 year‑old about what not to talk about publicly”, because of those very concerns. That is the challenge. When you introduce security questions in an educational setting, the bonds of trust start to break down, and the extent to which parents and teachers feel confident in their capacity to engage with such issues starts to erode. The more we can do to embolden teachers and parents to engage with these very difficult questions, the better.

More generally on Prevent, it has a chequered history. I still believe that the intention of the policy is a good one. Nobody wants to see people become involved in violent extremism and hurt others. The policy has been hampered by mistakes: weak conceptualisation, poor focus and inappropriate targeting, particularly of communities. I am sure these things are not new to you. There are some good news stories that we do not hear much about. When it comes to practice on the ground on the part of grass-roots actors, youth workers, community activists, police officers and probation officers, although there are examples of poor practice—there is no doubt about that, and mistakes are made in the referral process—there are also some good news stories. There are dedicated people doing some really important work. The challenge is that nobody wants to hear about their work. Equally, because of the security framework that sits around Prevent, nobody wants to tell those stories, because as soon as you are identified as being associated with Prevent it becomes impossible to tell the stories in a way that enables you to continue your work. Essentially, the community loses faith in you as an independent interlocutor within particular spaces. It is exceptionally difficult. The history of Prevent is going to make it hard for it to move forward in a productive way, but that is not to say that its intention is not appropriate and that there are not important things being done at the grass roots that could be celebrated more.

Q128 Baroness McIntosh of Hudnall: What has just been said is such an important thing. You say that the policy has its heart in the right place, to paraphrase, but what would you recommend to take forward that policy, or a policy, so that it did not have those slightly toxic connotations? In particular, given what you said about bringing the notion of security into an educational context, can we reverse that, or in your view has it got to the point where we cannot easily do that?

Dr Sarah Marsden: It has become extremely difficult. People have become incredibly entrenched on both sides of the debate. I am trying to think about how the positions in the anti-Prevent lobby could be assuaged or addressed. I do not think it will be at all easy to depoliticise Prevent. One way is to focus on the question of safeguarding and incorporate the strategy more cohesively within the safeguarding framework, so that it is in the list of things that parents and teachers need to take account of, along with child sexual exploitation and other things. More generally, it will be very difficult to rebrand it or rethink it in a meaningful way. I am not sure I have the answer to that question.

The Chairman: Do you, Dr Awan?

Dr Akil Awan: I am not sure I have an answer to that question. I would be making a lot of money if I did. I agree with Sarah in the sense that Prevent starts from a laudable place, but it has become a toxic brand, particularly among the communities that it really needs to have on board for it to work. That is fundamental. Irrespective of the successes, if you do not have those communities on board—the National Union of Students and many of the major Muslim organisations—I cannot see a way forward.

If I was to give any advice, you need a strong evidentiary basis for whatever you do next. That is fundamental. One of the issues is the focus on ideology being central to violent extremism. Pretty much every study that has been conducted has disproved that or at least says that ideology is not the starting point. People appropriate and imbibe ideologies later on, but if you are talking about how—not why—people become radicalised, peer networks, social networks and spaces and all those sorts of things are far more important.

If we are talking about the narratives of extremist groups, narratives in and of themselves have no power; they have power only when they intersect with real-world issues or circumstances. Someone mentioned France or Belgium earlier. I have done a number of studies on people from France and Belgium who joined ISIS. Some of the propaganda and messages that come from ISIS target the lived experiences of those people. France is a particularly good case study. There is a general antipathy towards Muslims or Islam. There is growing xenophobia, whether that is seen in the rise of the far right, or sartorial restrictions on Muslim women’s dress—the burkini ban and that sort of thing. It is really about belonging. If groups are offering a kind of Utopian state with welcoming arms, it can be particularly appealing to young people.

Conversely, if you look at socioeconomic marginalisation, in France Muslims make up about 8% of the population but 70% of the prison population. I can give you examples of a couple of ISIS social media campaigns. One is: why be a loser when you can be a martyr? If you are appealing to someone who is potentially socially, politically and economically marginalised, that is a very powerful message. I am not sure why contesting that message without doing anything to change the structural conditions in which that individual finds himself would have any power at all. Why would it?

We need to think hard about some of the reasons why people become involved in those sorts of groups but we also need to take a look at the evidentiary basis. We do not need to reinvent the wheel. We have been dealing with political violence for a very long time. Before the Islamist threat, we had the Troubles. We have a whole swathe of literature that deals with some of the very rich reasons why people are involved in these groups, and how they disengage afterwards. We should be looking at that rather than basing things on anecdotal evidence, putting forward policies and realising in hindsight that perhaps they were not the best thing.

Lord Sheikh: I want to raise an issue regarding Prevent. I have a strong feeling, talking to the community at every level, that it needs revision. You describe it rightly as toxic. In your studies, how much does the fact that the West is involved in Syria, Iraq or Afghanistan contribute to people being radicalised? It is not all ideology. What is your feeling about overseas policy?

Dr Akil Awan: To clarify, is your question about the role of foreign policy in radicalisation?

Lord Sheikh: Yes. People see what is happening in Syria or Iraq. How much does it contribute to radicalisation? As you know, the invasion of Iraq went down very badly. How much does that influence people?

Dr Akil Awan: It is very hard to measure something like that. We have to be careful that what we say and what we do match, and that there is no cognitive dissonance between our aspirations and ideals and how we act in the real world. For example, if we engage in military intervention in Iraq and that leads directly to the rise of ISIS, as it did, those two things are related and people recognise that. If we decide, for example, to cut our international development aid, but continue to arm pretty nasty regimes around the world, there is a gap between aspiration and reality. Young people in particular are quite savvy about recognising some of those disconnects. Groups prey on that as well. Groups are very keen to play that out in any number of ways as being an attack on Islam. I mentioned the Crusades earlier. The jihadists of the world talk about the current conflicts as the Zionist crusader alliance. A problematic foreign policy is very easy to package as part of the extremist narrative. No doubt.

Lord Sheikh: But it does influence people.

Dr Akil Awan: It is hard to measure, but yes, of course.

The Chairman: Thank you both. I for one have learned an awful lot from you two. We are very grateful to you. Thank you very much for coming in. It has been an excellent session.