15
Examination of witnesses
Sam Wise, Nancy Hey and Alfred Malmros.
Q92 The Chair: Good morning, welcome and thank you, all three of you, for coming to help us out with this really important and challenging inquiry into national resilience and preparedness.
Before we get on to the questions, I remind you that this is a public session that is being recorded live. A transcript will be sent to you in a couple of days so that you can correct any minor mistakes. If anything occurs to you that you should have said but did not, we always welcome any supplementary evidence in writing that you feel would be helpful. When you give your answer to the first question, it would be helpful if you could briefly introduce yourself for the record. As you know, we have a number of questions for you; I will start off with the first one.
If we were looking to identify a baseline of information on which to build about public awareness and preparedness in the event of severe national or international risks, do you have any data that gives us that level of awareness of preparedness at the moment? How does that vary among different segments of the population, such as between the young and the old, males and females, or people who live in the devolved Administrations, for example?
Nancy Hey: Thank you, Chair. I am the director of evidence and insight at Lloyd’s Register Foundation’s Global Safety Evidence Centre. In partnership with Gallup, we run the World Risk Poll, which has a resilience index in it. It is a global survey that we have done every two years since 2019, pre pandemic. Today I will share some of the latest results, which will be released later this year.
The survey is helping us understand people’s understanding of risk and their experience of harm globally, as they perceive it. It also tells us how the UK polls, and how it does against 140 other countries, many of which we do not have official data for, so this is really rich information. We can also look at it alongside UK surveys such as the 2025 Cabinet Office survey on resilience, which has some similar questions because we helped inform it. We can also combine it with other datasets. For example, with the global Resilient Cities Network, we have been combining the resilience index with other commonly used metrics on resilience.
From that information we can see that UK resilience is remarkably stable. The UK is one of 10 countries in the world that have not really moved over the four waves of the poll over eight years, but it is not particularly high on the list. That is against a backdrop of globally declining resilience, so it is not entirely a bad place to be, but, within northern and western Europe, the UK ranks about 15th out of 17 countries on our resilience index, ahead only of France and Latvia. It is stable but low.
Community resilience is a relative strength but with a strong reliance on government. We can see that individual planning is relatively low, although it is increasing gradually. One-third of people have a plan for what to do in an emergency that is known by all those over 10 years old in the household, and about 40% of people in the UK are confident that they can protect themselves in an emergency. That is about average for Europe. It is fairly consistent with our past exposure level to disasters. People are getting it about right for what they have previously experienced, but awareness of risks is growing.
Two-thirds of people are concerned about risks and, for the first time ever, we are seeing concern about war being a spontaneously given response, by about 7% of people, making that the fourth-biggest concern, behind road accidents, their health and the economy. This is the first time we have seen that in the eight years of the dataset. The risk backdrop is changing, but people’s muscle memory is not the same. About 27% of people have a plan, which shows that preparedness is gradually increasing. It improves after major incidents, particularly disasters of severe weather, but we rank about 75th out of 140 countries globally in terms of having an individual plan.
Sam Wise: I am the chief strategy officer at Saatchi & Saatchi, the leading communications agency. I am a communications strategist. My job means—this is often undervalued—thinking at the outset about what communications can achieve. You cannot make up down or black white; you are mainly in the business of accentuating, emphasising and pushing things in a direction, then working out what the right audiences are to effect change and where to go in order to reach those people. Finally, we work out what to create in order to make those impacts, and then create them. That is the work of a communications strategist, and it is probably the way I can best serve the committee.
When we received this amazingly interesting briefing, our starting point was that it feels like some things have changed in the past 10 years in the way that people communicate and trust certain sources of information. I think we can all feel that, but I wanted to come to the committee with a bit more than feelings, so I reached out to my friend who runs a company called Early Studies, which has research techniques to uncover things that people do not often show, allowing us to investigate things in a bit more detail. I will try to add to what has just been discussed with some things that we have looked at in relation to resilience and what we have noticed.
First, people expect crisis now. We saw some quite profound changes. In 2019, asked how likely they thought it was that the UK would face a serious national crisis in the next 10 years, about 30% of people said that was certain or very likely. Projecting out to 2031, that number is up to about 70%, so there is a big sense that it is coming.
Secondly, I draw your attention to the education in crisis we all got during Covid. If you ask people, “When you think about a major crisis hitting the UK, what are the things that people around you would worry about the most?”, there are some baseline factors that you would expect to stay the same, and they do: their personal physical safety and the safety of their family and immediate community. They are the top things, and they come up quite high, but we can see an increase in people’s nuanced understanding of the impact of crisis. Short-term disruption is quite an obvious factor, but longer-term economic damage to their finances or job is a second-order factor, and, even more nuanced is the breakdown of social order, the collapse of public services and even the loss of cultural stability and national identity. People have quite a nuanced understanding of how this is going to affect them.
The third point I want to raise, which is quite stark, is a collapse in people’s belief that central institutions will be well placed to handle a crisis. I think we can feel that, but I can put a few numbers on it. When asked in 2019 how confident people were that the country would respond effectively if such a crisis came, about 65% were either very confident or cautiously confident. That was a majority of people—about two thirds of people thought that we would be okay.
Now it is down to 28%, but only 3.5% of people are very confident that the institutions would hold. When we are doing a communications strategy, we look for a tension such as that, where one thing is going up and the other thing is going down. That creates a need to do something different. That is at the heart of what I would like to talk to you about today and some of the implications. It will inform my answers to further questions.
Alfred Malmros: As Sam explained, I run a company called Early Studies. Before that, I spent a decade at Google. Five of those years were at a geopolitical technology incubator called Jigsaw, where we focused internally, looking at our own data and what technology solutions provide, mainly in the face of conflict and geopolitical conflicts. We were central to most of the investigation in the technology sphere back then with mis- and disinformation and the impact technology had on those elections. I would add that that has also changed; there is a sort of human wind chill factor where everyone now believes we are currently in a crisis. With the exposure on social media of news from all around the world every second, there is a permanent feeling that we are in a crisis. That is something that is changing as well. When something happens, it is well worth noting that—how do we make sure that people understand that it is happening?
Q93 Baroness Hunter of Auchenreoch: Thank you for coming. I am always overwhelmed that we are able to have such expertise. I was struck by what Sam Wise was saying about Covid, the pandemic and how to do effective communications campaigns. I remember that it was very effective. It was “Stay at home, protect the NHS, save lives”. It was a pithy phrase. Now, we have “See It. Say It. Sorted”, which I think everybody understands. First, what do you think is the pithy way to do it? Is using these short phrases the best way? Secondly, should we use celebrities to make the communications more effective and more trustworthy? We met some young people last week and the name they came up with as a trustworthy figure was David Attenborough.
Sam Wise: Frustratingly, often the answers to difficult questions tend to be “and” answers rather than “or” answers. There is obviously a role for centralised communications, but we can see a big drop-off in how much people listen to those. The bone structure of an effective communications campaign for behaviour change is still the same: you need specific, achievable actions for people to take which are memorable and that can stick in their head, hence the slogans. It needs to be linked to something that they already value or worry about, and then it needs to be carried by a messenger they trust. The problem is the change in where that trust lies. In all the data that we look at, everything that is centralised—whether it is national broadcasters or the PM—is declining and everything which is closer to you is increasing.
I can pull out a couple of interesting stats. When something serious is happening, who do the people around you turn to first for information? This is just about access to the information, not credibility. National television has gone down from 82.6% to 50.8%. This is something that we have seen and it has been happening for a while. Change happens gradually and then all at once. Family and friends via WhatsApp or messaging has gone up from 33% to almost 50%.
We try to think about communications in four ways. There are paid communications, which are campaigns that you put out, such as ad campaigns. There are earned communications, such as PR, which is somebody else talking about you or doing an interview. There are shared communications, which are designed to live in a social space, and you must accept that you cannot completely control them. You are looking for people to interact with them. Lastly, there are owned ones, which is your website. Paid and owned communications are things that you can control, whereas shared and earned are things that you must—and it is very hard for people and businesses to do—relinquish a certain amount of control in order to set them free into the environment. It is a layering of these kinds of things. Getting the indirect communication strategy going is the major thing that I think needs to be added.
The Chair: Let us hear from Mr Malmos, who is an expert on human behaviour.
Alfred Malmros: Yes, slogans are incredibly effective if you have an enormous amount of resources to put behind them. It is everyone’s dream to create a memorable slogan. But, with celebrities, that phenomenon is changing. Today we are just coming to terms with what an influencer is and realising that these are the new celebrities. But, in a lot of our research with young people, we are seeing that they are becoming closer and closer to them, and it is becoming less centralised. We are no longer talking about influencers with half a million followers but niche influencers with 5,000 followers who are very relevant to young people. Influence from someone will always be key, but everything is gradually being more decentralised.
Nancy Hey: Briefly, the UK has much higher than average trust in local news and weather for information about disasters; the figures are 85% and 70% here versus 50% globally, so that is quite a big difference. About 10% of people will trust famous figures or community leaders in that way, compared with a third globally. The level of trust is different, and they might get news from different places. There is quite a difference in terms of where people get information they trust from. Text message is very clearly somewhere that young people are more likely to get information from, but not the very poorest. These different channels vary for different people. We are seeing much more fragmentation.
The Chair: Baroness Curran and Lord Oates have supplementary questions.
Q94 Baroness Curran: This is fascinating and helpful. On that point about people hearing different voices and how they respond to them, can I ask two questions? First, is there a gender analysis on this? Do women and men respond differently to different influencers? For example, how do women respond to men, and what are the issues around that? Secondly, when we met the young people last week, a number of them said to us that they did not really tune in to the word “resilience”. They saw resilience as something that the Government would do, not something that they would see as their responsibility. Maybe we are not using the right language. Is there any research about the words that we should use and how significant is that in terms of how you get people to hear what you are saying to them?
Q95 Lord Oates: You mentioned trust in local news and local weather. Is there a difference between levels of trust in national government and in local government and, if there is, is it significant?
Nancy Hey: Globally, there was not a great deal of difference in age and income, particularly on who they trust. But that may be more different if you look under the bonnet in that information.
Baroness Curran: Is there a difference in gender?
Nancy Hey: There was not a huge difference, but you are right to say that in the UK, people are much more likely to see the resilience as something that the institutions of national and local government do. There was not a great deal of difference in terms of national and local government, but local government was slightly more trusted on preparedness. We have questions about how confident people are in these institutions’ preparedness. It is interesting that people are more likely to think that it is the Government’s job, but do not necessarily feel particularly confident in their level of preparedness. This is quite difficult for the UK, where our national critical infrastructure has a fragmented delivery system. There is a risk around that ability to directly control it as well. That is a challenge.
The words people use and understand in terms of resilience are definitely something that people do not connect with. One of the biggest challenges in the UK is particularly where we have people who do not know what to do in an emergency. There is no individual muscle memory. There is a community one. We know where to volunteer and we saw that in the pandemic. But as individuals, we do not know. It is not necessarily a motivation issue. It is an understanding of practically what you do. It is gradually creeping up but there is something quite particular about what you need to do in response to that concern, because we do not have that.
Sam Wise: When we talk about the central tension of a belief that crisis is coming and a lack of confidence in central institutions, it is interesting that that holds across most of the audience groups we looked at. We need to think about splitting up audiences in different ways, not as an external force in categorising them. Whether in the private sector or anywhere, you get more meaningful audience differences when you start splitting people by their values. A good proxy for that might be political affiliation. When we look at our data, the belief that the centre is not going to be able to hold up is accentuated the further away you are from the centre, as you would expect. That holds true across different levels of wealth, for example, and is broadly similar across men and women.
Q96 Baroness Helic: Thank you for your evidence. That belief that you have just described, would that hold in the times of the actual crisis? We are all speculating. We are all expressing our views and lack of trust in institutions and central government. It is very easy when it is theoretical, but when you are actually pressed, you do not go to your influencer. You go to your local weather, I think.
Sam Wise: That is a great question and one that we talked about in preparation for this. You have to be clear about the different stages of communications, because when bombs fall from the sky, everyone will still go to the underground. So the crisis itself broadly takes care of itself because you have the stimulus of seriousness. The real communications strategy where thinking is required is in how you prepare people for it when something is not there. Then it is how you make people adhere to it afterwards, when there is all kinds of different information that people can cherry-pick to support whatever course of action they like. The three years of Covid was a lesson in that. People adhered to things very clearly at peak moments but it was on either side that it became harder. So most of what we are talking about is in the absence of the direct stimulus of the crisis happening; it is on either side.
Q97 Baroness Mobarik: How can the Government encourage the public to be more self-reliant and resilient, including through the promotion of civic values and drawing attention to domestic and international pressures? That is quite current in terms of civic values.
Alfred Malmros: It is a fantastic question. On the encouragement of self-reliance and resilience, we have seen enormous shifts, especially in the ages below 30, with an incredible increase in the idea, value and virtue of agency. Having grown up in a constant state of at least being told we are in different crises, people have become enormously more self-reliant and a lot less expectant of help, essentially, but take it as a great value and virtue. Intentionality is one of the guiding principles of the next two young generations. When we have done research on the decline in drinking, it all comes back to wanting to be in control, in charge of themselves, having agency and being prepared for whatever hits them.
Nancy Hey: I understand civic values to be about societal and community resilience. It is one of the UK’s relative strengths globally. It has held stable over the eight years that we have polled. In the UK, about 80% of people think most of their neighbours care about them and their well-being. That is very high compared with the global median of under 70%. We are about 20th or 21st in the world at that. So, despite what we are seeing, that it is a relatively strong side of it. That is how I would understand civic value. That individual one is the gap. For me, part of it is that people just do not know. There is no muscle memory there. For example, in the floods in Spain recently, people did what they thought was the advice the previous time, which was not very recent. So they all went to the basement, which was exactly the opposite of what they should have done. This is what I mean by a muscle memory.
We have a memory of volunteering. We know how to help out in our communities. We saw that in the pandemic. We do not have that on the individual side in the same way. That is something we have to teach a little bit, because we have it about right. But there is a lot of science and practice around risk communication and helping people prepare. We fund something called the Risk Know-How project that has a lot of experience about how you help people communicate that. People are making complex risk decisions about whether they go to work or not. They are dealing with multi-hazard risks in their own individual circumstances. That multi-hazard risk is really important. For example, in the recent hurricane in Jamaica they were worried about flooding, so they moved everyone to the top floor of the hospital but then the roof blew off the hospital. What do you do in those circumstances where you have multiple hazards happening at once? That science is expanding but we need to learn from it fast.
Sam Wise: This is a kind of bad news/good news story. The bad news is that there is, as I have referenced already, a lack of faith—certainly a decline in it—in direct government communications. When we ask people, what they think is driving government communications, the sense that the Government are genuinely trying to keep people safe is declining, from about 52% to 31%, whereas the sense that what is driving it is the desire to protect themselves from blame and legal implications has increased from 43% to 57%. What we can do is try to embrace indirect forms of influence. That would be my answer.
Alfred touched on something that is a mega trend that we can see in all walks of life. Many private companies are trying to cater to this, whereby people are taking the organisations that they used to rely on, whether in terms of health, retirement or education—things that used to be done in a centralised institutional way—people are taking responsibility for themselves because they think that they can handle the complexity, often empowered by technology, and that they can take that on themselves. You are seeing people challenging minor legal disputes using ChatGPT. You are seeing family crises with a loved one in hospital using Claude to try and interpret the medical side. The doctor comes in and you know what three potential diagnoses are coming up.
When handled right, this transfer of responsibility from the centre to individuals is felt as very empowering. When handled poorly, it feels like you are dodging responsibility. But it is incredibly powerful how much we are seeing this. If we ask, “In which areas do the people around you feel they are increasingly making their own decisions?”, the figures are so stark. There has been an increase from 30% to 61% in health and medical decisions. If you sit around a dinner table now, there are more diets than people. What we eat is different. How we exercise is different. What we think we need from a health perspective is different. But even in things that you would have thought were completely collective, such as education and what your children learn, only one-fifth—21%—of people thought that that was something they should decide themselves. But for a long time now I have been sitting at dinner tables with people who are clear about what their kids should be doing in the future.
About 50% of people now think that that is something they should try to take into their own hands. The companies that are helping people doing that are doing really well, and we work with several of them. The way to influence people’s resilience is to help foster their own resilience and empower them to do it themselves, without looking like you are dodging the responsibility, of course.
The Chair: Lord Peach has the next question, which develops this theme.
Q98 Lord Peach: Thank you all very much for coming. Sticking with the theme, the big question for all three of you—I do not mind in which order you answer—is: should the Government be more honest and truthful with people about the threats and risks that the country faces? There are those who do not want to push that too hard because they think it would frighten people, whereas I think this morning’s suggestions say otherwise.
My second question follows the brief discussion about influence. Some of our Scandinavian friends give everybody a brochure about what to do when the war comes, and the UK has been very reluctant to do that. I do not know whether any of you have seen these brochures, but they are very good. They also do one for industry. I wonder what you think about that, as a secondary question, but the big question is whether we should encourage the Government to say, “These are the realities. These are the risks and threats we face”.
Sam Wise: That is a big question, and I would not hope to be able to give the perfect answer or a completely black or white one. People already do not believe what the Government are saying to them. They think that they are motivated by other motivations than just to serve them. What people are looking for is less of a sense of narrative. People are becoming quite sophisticated at spotting narrative and want less of a sense of the analysis put on top and more of the facts underneath. Some more truth would probably be useful. If you held my hand to the flame, that is the way I would probably go.
Nancy Hey: We know from Risk Know-How information that if you have more information and knowledge about an issue, you also need to have something that you can do about it alongside that, so that you do not feel hopeless. Knowing that there are things you can do and a pathway to get there are the ingredients for hope, including practical, grounded hope. Information that says what you can do is quite useful.
Information is important. You talked about the very simple messages about what people can do, which are important, particularly in a crisis, but there is a space for trusting people with more detailed information. People are not in individual buckets. They are not just members of the public; they also run their business, help in the local community or are on the residents’ association. They are not in a single bucket, so what we need are information flows—linking social capital between the different networks people have, which is slightly more detailed information than crisis information, but it is more trusted as well.
Alfred Malmros: Full disclosure: I am Swedish. I remember that, when Covid came to Sweden, the first thing they went out and said was, “This is the plan we devised for this situation three years ago, and we are going to stick with it”. That built an enormous amount of trust because the message was that there was already a plan, not that there was now a plan. Showing that you have planned for something is crucial for trust-building. It was very effective in Sweden, where people maintained trust in the plan for a long time because it was not a reactive plan.
On truth and information, if a new crisis came, the first thing a lot of people would do is find all available information online on the subject, put it into ChatGPT, and ask, “What does this mean for me?” The real questions will be: who puts out that information, how much do they put out and who owns it? That will be crucial when it comes to transparency, truth and plans. It is not so much about offering the recommendation but offering up the data on which the recommendation was made.
Baroness Helic: In that context, do you think that all government information pages should be made scrapable? I do not know whether that is the right word to use.
Alfred Malmros: No. What amount to make transparent is not my area of expertise. This tacks on to other themes such as disinformation and misinformation, but there will be an advantage to whoever puts out the largest amount of information that you can feed into an LLM to get answers back. That is an enormously important question for national resilience.
Q99 Lord Oates: Thank you so much for your fascinating insights. Could you talk a bit more about how you see the content of messaging on preparedness and resilience being tailored to different audiences? I am interested in both the content and who delivers it, which you have touched on a bit already.
Sam Wise: It is absolutely the right instinct that messages need to be tailored to different groups. We need a slightly more sophisticated way of looking at what those groups are, which I touched on. As a wheelchair user, I would be classed in a vulnerable group, but within that group there are people who have to drive themselves around using tennis balls in electric wheelchairs, who require a lot of help, there are very elderly people and there are people who would consider themselves active. This is not a group that I ever applied to be a member of; in fact, the only thing that the disabled community can really agree on is that blue badge parking abuse should be punished by death.
There is a way of defining a group externally, and a way of defining it internally. It is much more helpful to look at what people’s shared values are and what motivates them. That is the starting point for mapping what those audiences of influence are. We have to divorce this from the sense of a geographical map and start thinking about a digital map of communities, who people cluster around, who are the important voices, and which communities are the most important. That is step one.
We talked about the paid, owned and shared media. For example, very wealthy people tend to be biased towards owned media. They do not want to hear what somebody else says; they will go straight to the site and get the information for themselves. They have a different spread of influence. You need a sense of not just who the audiences are but what the places are and who are the people to influence them. It is a job of work to do that, but it is a job of national preparedness so that you can pull those levers.
You talked about Sweden; I think of its K-companies. You need something similar for a communications network of influence because, as Alfred said, you want to have the domain authority and to be the one releasing the information that will then influence all these other people to have a coherent message, even if it is not the same message. What we are seeing is that, what people now want in order to believe something is not the same message 15 times, which is the classic communications strategy, but messages from different sources which resonate with each other. The phrase, “Do your own research”, means that you want to see something in a number of places so that you can then triangulate that it is kind of true. What you are trying to do is influence and seed stories that can happen in different places. There is an expression we use a lot: if you want to gain influence, you have to surrender control. Those two things are counter to each other. It is very hard for institutions to do that, but it is necessary.
Nancy Hey: We have talked a bit about where people trust in things, but, every day, people make complex decisions in very complex environments for their own benefit. That is what I referred to in the Risk Know-How information. People are making these decisions, sometimes about information and sometimes about who they trust, and you can see in the science on trust that you can have the same information from different messengers with very different outcomes, so the content and the audience definitely matter.
The group of people whose level of preparedness I was particularly concerned about were the 15 to 25 year-olds, who were least likely to have a plan and probably quite unlikely to trust authority figures. I am particularly concerned about that group but also the people who do not necessarily have access to information. They are just above the very poorest, so they are not in the target group that are particularly vulnerable. They are just above that and do not always get the information they need to make the best decisions for them, often making very complicated decisions about whether they go out to work because they need the money.
It is important to think very carefully about who people are in lots of different spheres of their lives; they are not just one thing.
Alfred Malmros: I do not have anything to add to those brilliant responses.
Q100 Lord Oates: Nancy, in the previous answer, you said that there is very little difference in the public’s perception between the preparedness of local government and national government. But you also gave evidence about the degree to which the public think that the Government care a lot about their well-being; that is not broken down into national and local. Is there any specific data on that or does that not exist?
Nancy Hey: I think that is different. People do not necessarily distinguish between national and local that much. We have data on whether you have confidence in your Government, whether you are confident that they have prepared and whether you think that they can be trusted with your well-being. It was interesting what we were saying about the evidence on trust and plans from business, charities and states. Trust comes from having a plan with somebody’s well-being in mind and people trust that that happens until it does not; then, there are systems in place to monitor that plan and adjust it in the light of events. That is what we saw, certainly in businesses that were able to be resilient in the pandemic. In many cases, those businesses, fairly serendipitously, happened to have a plan and systems in place. That then enabled the leadership to be authentic and trusted, because, even if they got the plan wrong, they had systems in place to monitor and change it.
The Chair: Baroness Northover has a question about the flipside of information.
Baroness Northover: There are so many things that have emerged from what you are saying. It has been fascinating. Thank you. If possible, can some of Mr Wise’s charts be shared with us?
Sam Wise: Yes, I would love to share them.
Q101 Baroness Northover: The information in them is incredibly important for us. I am thinking about how young people are the least likely to have a plan, least likely to trust and do not have the best information. But they are going to get older; it is not that that is always going to be the case with young people. That is the generation that is coming through, so that is one of the challenges that we are facing that we need to be addressing. Also, in the Covid pandemic, it was when much more detailed information was coming out from Professor Whitty and Professor Vallance, rather than the original short messages, that there was very thorough engagement. This leads me to my question about disinformation and misinformation in terms of elections. This is obviously an incredibly worrying area, and it relates to trust and everything that you have been talking about. Do you have any suggestions for how those risks could be mitigated?
Sam Wise: Perhaps a little controversially, something we talk about is that one person’s misinformation is another person’s activism. Even in mainstream media, the same event can be reported in the Guardian and the Daily Mail as two completely different versions. There has to be a distinction between disinformation, which is malicious, and what can loosely be called misinformation, where people are putting out things which are not necessarily true. We should be trying to bring people closer, not to try to control them, but to be the source of information. You have a power over them in the information you are able to give them. If you are able to relinquish some control over what their exact narrative is, you can start to make things more coherent over time.
There is another part of it, which Alfred is really well placed to speak to, which is that there is also a danger of the confidence—which I am sure we have all encountered in our lives—that people who do not really know anything about a subject get very quickly, with the power of LLMs. It is a sort of artificial intelligence confidence, and that is one of the things that really heightens the importance of understanding the sources that are driving the recommendations of LLMs. It is important to make sure that those tie back to things which can have really accurate information and where the domain authority, the speed of update and the “liveness” is all influenced by the Government directly.
The Chair: Would either of you like to come in on that?
Nancy Hey: The only data I have is from 2019, so pre pandemic. UK citizens generally were more concerned than the global population about false information when using social media—about 60% in the UK and a slightly lower figure globally. Age and income did not matter; people were concerned regardless. We have not polled that since, so maybe we should.
Alfred Malmros: It is fairly anecdotal now, but when we look at misinformation and disinformation in the US election in 2016, most of the headlines were about the amount of disinformation. We actually found little of that in data from YouTube, Google, Facebook and Twitter. But there was an enormous amount of misinformation in bubbles, which was brushed away. There were echo chambers where everyone was using numbers that they had found somewhere that they believed to be true to convince each other even more that they all agreed on something. That was the first time that we really saw how these incredible separate and distant worlds were created, by those who had completely different ideas of what the world looked like. Since then, we have had a generation that has grown up online and that does not know a time where there was a central source of information; they have grown up with 20. They know full well that they should check out a right-wing opinion and a left-wing opinion, and the truth is somewhere in the middle.
Back then we often used an anecdote about Finland. How does it have virtually zero disinformation, given that it borders Russia? Finnish students learn critical thinking in the national curriculum. From the start of school, they are taught to question the information they receive and create their own opinions. It might sound absurd, but this critical thinking native generation has had so much information, it has been forced to build up tools for managing it. Most of the stories we hear are about people who do not manage, but I think they are a very small minority. When we have done research recently about people’s worries about misinformation, most of it is from young people about their parents, who have not grown up with a stream of voices from everywhere. Meanwhile, the parents think that they are the ones who need to be worried about the children. Younger people are taking care of it.
Going back to LLMs—ChatGPT, Claude, et cetera—we will also start to see a couple of interesting phenomena coming up there. In three years’ time, maybe, we will be talking about misgeneration or disgeneration; that is, LLMs creating information either from a nefarious purpose—models that come from authoritarian regimes, for example—or from various different biases. The only thing they have in common is that they need information. They do not have it themselves; they generate it from existing information. That goes back to the role of information—what information the Government provide and how it is structured. That is going to be crucial.
The Chair: Lady Winterton, did you have something to add on elections?
Q102 Baroness Winterton of Doncaster: Thank you for the presentation so far; it is very interesting. I want to ask you for your reflections on some international aspects of misinformation and disinformation. I have just come back from Armenia, where, with the OSCE, the Organization for Security and Co-operation in Europe, I was a short-term observer in the election. Previously, I did that in Moldova. In both countries, but particularly Armenia, where there were very polarised views and vigorous debate between the very pro-Russia population and those that are pro EU, some of the debate was very polarising. I suspect that reflected what politicians themselves had said about what would happen if they moved away from Russia and towards the EU, Azerbaijan or Turkey. I wonder whether you have any reflections on the tactics that have been used by Russia in some of these places.
Interestingly, the first short-term report from the OSCE said that the deliberate misinformation on social media did not have that much of an impact. Perhaps that is because it was particularly aimed at young people, who, as you say, could distinguish between what was just political dialogue and what was real misinformation. Do you have any reflections on that or experience of it? Have you ever looked at the reports?
Sam Wise: I am not particularly qualified to speak about the specifics of that, but, as Alfred mentioned, looking at young people’s ability to spot gen AI or generated pictures, you see that there is a level of education that comes with just growing up with that. There is one nice piece of positive information here. Looking at what kind of information people act on during a serious crisis, you find that those who can verify for themselves from multiple sources has jumped from about 30% to one of the biggest numbers. That sense that you should be sceptical and start from a perspective of critical thinking is a mark of intelligence. Being easily led is seen as the opposite. That is cause for optimism.
Nancy Hey: I do not have a view on the information side, but I see that, for the first time in the eight years that we have polled, politics has emerged as a top-of-mind risk to people’s safety, but only in North America.
The Chair: We have only a few more minutes, so I will ask Lord Marland to ask the winding-up question.
Q103 Lord Marland: The big question is: what recommendation would you make to the Government if you had every opportunity to change things?
Sam Wise: You need a supplementary indirect communications strategy, which is about how you influence other people to talk about how to handle crisis. That needs to be in place beforehand. For that to be effective, you need to map the landscape of influence and communities in a digital space, and then you need to keep track of what the key nodes or voices are within those. That would set the country up to be able to handle a crisis, complementary to central communications. That is the missing piece. At the moment, that stuff is happening—people are going there for information, and it is the Wild West; there is no indirect influence. Having an indirect strategy of influence is the main thing that I would suggest.
Nancy Hey: This House having this civic conversation is exactly what we should be doing. Community resilience is a strength in the UK, but it works differently from other sources of delivery. It is much more fragmented, and it is motivated intrinsically, but we should build on that strength. We should help people have a plan because they do not know, since they have had no need for it in the past. There is a strong reliance on central government, but we have a fragmented national infrastructure delivery system, so information sharing and links are really important. I would probably recommend something like a chief learning officer to pick up all this learning across different sectors so that we can keep learning as we go.
Alfred Malmros: On Sam Wise’s point, when something happens, people will open up WhatsApp; I do not think they will go to bbc.co.uk. In that WhatsApp channel, someone will be seen as the authoritative voice. That person will have checked out what is going on with ChatGPT, which will have told them, “You are absolutely right. You’re brilliant. You are a military strategist”.
My recommendation is to be a purveyor of information and recommendations, not just recommendations; to have a very clear strategy for the information that is provided; and for that to be good information, because people will look for more and more information and fewer and fewer recommendations.
The Chair: Thank you very much indeed. That has given us a lot of food for thought, and we are very grateful for your advice and evidence. As I said earlier, if you think of anything else afterwards, please write to us. In the meantime, thank you very much again for your time.
