The threat of misinformation and disinformation

Speaker 1: Welcome to The World of Intelligence, a podcast for you to discover the latest analysis of global military and security trends within the open- source defense intelligence community. Now, onto the episode with your host, Harry Kemsley.

Harry Kemsley: Hello, and welcome to this edition of The World of Intelligence by Janes, with your host Harry Kemsley, and as usual, my co- host Sean Corbett. Hello, Sean.

Sean Corbett: Hi, Harry.

Harry Kemsley: So Sean, today no guests, which is unusual for us recently. But the reason I wanted us to get together is as we discussed before we started, let's revisit this topic of mis and disinformation. Now, the reason I'd like to discuss it is that I was recently merely a victim of disinformation. I opened an email that talked about the arrival of a package from the US by a certain courier company that said, as expected, as I was expecting that it would be needing customs payments made. And it took me to a page that looked astonishingly like the page of the said courier, and it asked me to put my bank details in. And as I started to put my bank details in, I suddenly realized I was about to use the wrong credit card or something, so I paused. And I went to try and go backwards in the process and the buttons didn't work. In fact, none of the buttons around the screen worked. That was what I suddenly realized was the clue that I was about to be scammed out of the cost of the customs fee. Now, the reason I mentioned it is because that's how pervasive it is, it's everywhere. And I'm somebody that would like to believe that I'm actually pretty good at spotting that sort of thing, and I pride myself in being involved in open- source intelligence. And yet there I was staring at what appeared to be a very credible and timely disinformation that was clearly driven by the need for money by somebody that was going to come out of my bank account. It didn't happen but, boy, was I close? I was one mouse click away from sending them a chunk of cash out of my bank account, which would have been really, really annoying.

Sean Corbett: So, that's a really good example actually, Harry, because people are going to no doubt say, " Why are you going through the subject again?" Well, firstly because it is incredibly topical, which I'm sure we'll come on to. The second thing, it pervades every aspect of society that we have in our Western liberal democracy. Whether that's the individual level that you thought, or at the national level where there are nefarious actors, some of who are state actors, some who are not, who are trying to shape and obscure the reality of what we see, and they're in many ways succeeding. This is a really important and difficult issue.

Harry Kemsley: Well, then let's ... We tried this once before and I'm going to try and use it again, this sort of framing questions up front. Let's just put some frame around this topic, so let me ask you a few questions. And the rules for this are, you can answer yes or no, one word answers. If you absolutely must, I have borrowed this from a lot of podcasts I listen to, and I quite like the technique. If you absolutely must, you can answer but only with one sentence.

Sean Corbett: That will be a struggle for me.

Harry Kemsley: All right, here we go. You up for this?

Sean Corbett: Yep.

Harry Kemsley: All right. Do you believe that disinformation and misinformation pose a significant threat to modern society?

Sean Corbett: Yes.

Harry Kemsley: That's interesting, very straightforward yes. Is there a clear distinction between disinformation and misinformation in terms of their impact and intent?

Sean Corbett: Yes.

Harry Kemsley: Can an OSINT analyst, an open- source intelligence analyst, always rely on traditional methods to identify and mitigate mis and disinformation?

Sean Corbett: Yes and no.

Harry Kemsley: As I expected. I knew the yes or nos would to be difficult. Okay, we'll definitely come back to that then. I do today want to try and spend some time talking about what can we do about it?

Sean Corbett: Yeah, absolutely.

Harry Kemsley: We'll get to that. Is social media the primary vehicle for a rapid spread of both mis and disinformation today?

Sean Corbett: Yes.

Harry Kemsley: Could unchecked disinformation, I've been very specific about the choice of word there for reasons that will become clear later. Could unchecked disinformation ultimately undermine the foundations of Western liberal democracy?

Sean Corbett: Absolutely, yes.

Harry Kemsley: Yes. Okay, that's pretty horrifying. But anyway, we'll get to it. All right. Then, because we've done this before, I feel this is potentially redundant, but for those that hadn't heard it, let's start, Sean, with the definition. It is important because I think as you and I will talk about in a moment, there's an awful lot of misunderstanding of mis and disinformation, and people are using the terms incorrectly. So how do we define mis and disinformation?

Sean Corbett: Yeah, it's a really important one, and clearly not enough people have been listening to our previous podcast because they're still getting it wrong. And even within, if you want to call it responsible mainstream media, I call it traditional media now because I don't think it is mainstream media, they're getting it wrong and they're confusing the two, and they're obfuscating and using them, juxtaposition with each other. So disinformation is deliberate activity by either a state or non- state actor. So in your case it could be a scammer to provide misleading or incorrect information with the intent to mislead, influence, or change behavior.

Harry Kemsley: So there's an intent aspect.

Sean Corbett: Absolutely. And then misinformation is incorrect information shared without an explicit intent to deceive, but often reflecting individual or group preconceived ideas. Now, as we discussed last time with Amil, there is a slightly gray area because if you're trying to impose your own perspective on somebody that could be misconstrued as either misinformation or disinformation because you're trying to influence people, so at that individual level. So there is that little gray zone there.

Harry Kemsley: But in that misinformation realm, where you're not intending necessarily, but you are supporting something, you are propagating it, you are supporting the misinformation by the fact you're supporting it in some way.

Sean Corbett: Indeed. And you could call it unconscious bias, actually. But it's a little bit more than that because there's an active way of doing it. " I believe this, I've seen something online, I support that, I'm going to share it with people."

Harry Kemsley: So if disinformation, inaudible say, if disinformation is deliberate and often state- sponsored, while misinformation is usually shared, but without a harmful intent necessarily, they both still cause potentially significant damage-

Sean Corbett: Yeah.

Harry Kemsley: Right? So, one of the things that has struck me about the mis and disinformation part of our lives that is such a big part of our life these days is how infrequently I encounter teams that are specifically set up to do something about it. In a sense of, " Okay, great, so there's a problem, what are we going to do about it?" And I look into the intelligence world and I don't see necessarily many examples. Maybe there are, this is my lack of understanding of teams trained specifically to identify and to debunk state- sponsored disinformation. Maybe that's something we should be doing. And similarly, in the misinformation realm, which is more pervasive arguably because it's probably more social- media driven, something that's going on that's actually talking about ground truth and actually equalizing the biases and so on they're showing in it.

Sean Corbett: Yeah, I think to a certain extent, certainly within the classified intelligence world, there is an awareness of that. And this is another difference between misinformation and disinformation. Disinformation requires a different approach to countering it, whether that is counter disinformation, or cybersecurity, or just shutting down those outlets. But also a way of actually letting the decision- maker know, " This is incorrect," and I think that's still fairly immature. Whereas, misinformation is a more societal thing and requires the education that is so sadly missing, although people are just starting to get into it now. So, there are a few small companies out there that do it and we spoke to Valent, with Amil last time.

Harry Kemsley: Yes, that's right.

Sean Corbett: But it's certainly not as industrial as it should be. And of course if we want to start talking about the big social media companies, they know exactly what's going on and choose not to anything about it.

Harry Kemsley: Yeah. And I guess that's one of those areas we'll probably touch on a bit further just now. Because if I go back 20 years before the serious advent of social media and the way we talk about it today, well before, I don't know that we were suffering so much then with this pervasive element of misinformation. There was certainly disinformation, but we would have called it propaganda there, I suspect. And around that the military and intelligence organizations of the world would've talked about operations in the information realm or Info Ops, PsyOps, perhaps a slightly more exquisite version of that. But we didn't really talk about misinformation in the way that we do today. So what's changed? Is it just social media? Is it technology has enabled us to have this conversation?

Sean Corbett: So, I think just to nuance that a little bit, so the disinformation thing has always been there, albeit as you said in a different term, whether that is propaganda or just deception. But that's from a military capability to a military capability. And I go all the way back to the right hook from the Marines in the first Gulf War where the Iraqis thought they were going to come in from Kuwait from the sea. That's legitimate and all the rest of it. But I think now what we're seeing is with the advent of social media and because no one's allowed to counter anybody else or you get canceled or all the rest of it, but it is about the proliferation of social media. And I've got an example here. So talking about the horrible situation that happened in Southport. And the rumor coming out that the suspect was an immigrant or an asylum seeker. There was an estimate that there were over 27 million impressions on social media to that effect within 24 hours. I mean that couldn't happen if it wasn't for literally the degree of that social media. And once it's out, it's the Wild West, it's uncontrolled, there is nothing you can do about it. And despite going, " It's not right, it's not true, it's not true," A, people aren't listening to that. And B, the attention span of news outlets, some of them, many of them and the individual is so short that you've moved on something else, so retracting stuff is hard. And it goes back to what I always say and I'll probably say every single podcast is normally the first version of what you hear is not right.

Harry Kemsley: Is not right. So you've got to dig deeper to get to the truth.

Sean Corbett: Yeah, and it's that balance between getting to the truth in sufficient timeliness to actually get it out there and everyone believes it, as opposed to, " Oh, no, this has happened and ..."

Harry Kemsley: But we've moved on, yeah. I looked at some news reporting around that video back in March'22 when Zelensky, the inaudible of Ukraine was thought to have been on the TV to say, " Down these tools, guys. We're going to surrender to the Russian forces." Which of course was quickly identified as fake news. When I looked at the news reporting around it, there wasn't much... And to be fair, by the way, I was looking at traditional news media rather than social media, so there's probably a big difference there. But in the news media, there appeared to be a very, very fast recognition that was a fake video. Now, having watched the video, it isn't very well done, so it's probably fairly easy to spot that doesn't look right. Something about the way he looks and the way he speaks, the way his face moves and things doesn't look right. But interestingly, it didn't get picked up as a real story as in, " Zelensky is surrendering." That didn't come through at all in the news that I saw in the trends. It came back very quickly, " Fake videos has been broadcast presumably by the Russian government." I didn't see the social media, as I've said, but I'm wondering whether we're starting to see an education, people are starting to realize this is a real problem and then that will start to permeate into the media environment. The media outlets will start to say, " Hang on a sec, before we publish this apparent news, let's just check its validity." Whereas, a period of time ago that was less the case. Or do you still think it's still a massive problem, we're barely even scratching the surface?

Sean Corbett: I think it's slightly optimistic, but I would say that the awareness has gone up because of the things like Southport. And it is starting to get into the narrative of the political world, although it's very unsophisticated. And my concern there is that people are now picking their understanding of the world and using information out there to support their... So, this is the second and third order of misinformation, disinformation, which is the echo chamber where you are basically relaying things that fit your worldview, which just accentuates it. And you'll see that in some of the AI algorithms, actually. So simple example, whenever you Google something that gets logged, " So, all right, they're interested in that." Or even more basic with your shopping trends, if you accept cookies and then suddenly you find that lots of adverts there are stuff that might appeal to you. Now, the counter to that is that I always reject the cookies, but I don't half I've get some strange adverts coming to me, things like for haircare products, which as you and I know I'm folically challenged, or makeup, which is each to their own, but it's not my thing. So, you can almost cut off your nose to spite your face. But the echo chamber, which some academics say isn't a real thing, it definitely is. I mean, even on... We're kind of old, so I guess if you use social media it's probably Facebook. There are people that you are not friends with just because of their life view, shall we say, and you basically filter them out so you don't get them.

Harry Kemsley: I do wonder, though, with the echo chamber thing, whether we understand how pervasive that is. I've used that word several times today, but it is a pervasive problem. I sat next to somebody recently, we both put in exactly the same search term into exactly the same search engine and the difference in our responses was astonishing. Absolutely astonishing. And actually, that point about cookies is quite an interesting one, because I know how annoying cookie things can be in terms of the saying, " Will you accept? " If you're saying no, I don't want to accept it, it then takes you through three or four steps.

Sean Corbett: Absolutely, yeah.

Harry Kemsley: Each time you change a page it asks you the same question, which is really trying to push you into accepting cookies. I don't think people necessarily fully appreciate that that's one of the areas of their lives they have control over in terms of the echo chamber they create around themselves. And as we've said I think in a previous podcast that echo chamber is not being created for mal intent, it's being created for commercial reasons.

Sean Corbett: Making money, yes.

Harry Kemsley: The marketing time they get by you seeing their adverts is what drives the advertising industry to do what they're doing.

Sean Corbett: But it's quite, sinister might be too strong a word, but it is because, exactly as you've said, you're driven to accepting them because you literally don't have enough time in this fast- paced society to say, " Right, I'm going to go through each one and reject them all." Now things have changed a little bit in terms of there has to be a button that says, " Reject them all," but you've still got to find that button so it's easier just to-

Harry Kemsley: I've also noticed recently they've started adding a different category, there's a thing called legitimate interest. If you get into that, these are companies that have, "a legitimate interest in what you're looking at." Says who?

Sean Corbett: Says them.

Harry Kemsley: Legitimate according to them.

Sean Corbett: Exactly.

Harry Kemsley: You've got to take that out. Well, so long as that's true, then the Wild West continues. So the misinformation problem continues to be a problem. I don't believe, by the way, we're ever going to get that genie back in the model. My view is that we now have to find a way of dealing with it. Now, a long time ago we talked about data literacy in the broadest and modern sense of that, the ability to discern for yourself what appears to be right and what appears to be something other than right, whether it's wrong or to be determined later. What I find interesting, and I guess my scam experience or near- scam experience is an example of that, is how easy it is to fall into the traps of what appears to be credible, certainly the advert, the information inaudible was timely and it was exactly on point in terms of the company, and so on. But I didn't question it. I took it at face value because it seemed to fit my understanding of the norm. And that's what worries me is that if we've been normalized into our echo chambers and we've become very, very clear about what we think the world looks like and how it works, we're strengthening that bias all the time. We're reinforcing it continuously.

Sean Corbett: We are, and this is a worry, I mean you and I have been in this business for a long time, so we do have that spider sense if you want to call it that. But it goes back again to what Amil was saying that if... So in your case, let's do that one first. Actually, you still had the presence of mind to check you went through, but-

Harry Kemsley: I'd like to believe that was the truth. The reality is it was only because I was about to use the wrong credit card that I needed to go back in the process, at which point the button didn't work. And then that's when my spidey sense went, " Why is that button not working? It's a credible company." And then I clicked on a few other links on the page and none of them worked. They were just literally on the page.

Sean Corbett: Okay, because the point I wanted to make was from an intelligence professional, we'll start to get into that in a moment, is that this is the difference between the art and the science of intelligence. If you are an experienced analyst, and you're used to certain sources, and you're used to understanding how good they are, we do need to talk about sources, then if something's not quite right, then you know to just check it and do that as you should be doing anyway, doing the assurance piece. We'll come back onto that. But I loved what Amil said in his, he said that, " Be wary of things that make you feel good." As an analyst, we've all been in the situation where you're really proud an assessment, there's a really important bit that's missing and you can't find it. So you go into second, third, fourth, fifth level of search to try and find it and then you come up with the, " That's it. That's the answer. Fantastic, I've got it." Now, that's the point you need to go, " Hang on a minute, this is too good." And so then you've got to try and check it particularly as that source is almost likely to be a single source because you haven't come across it before. So, that is where the art and the science mix and you've got to go on your experience. Now, I could count on myself here, and you've heard me again several times, talk about senior decision makers, particularly military ones who you've given your best assessment based on incomplete information, which is what intelligence is, and they've gone, " No, I don't agree that. I really don't agree with that based on my military judgment." What they really mean is that, "This doesn't fit my plan and it's not going to get me promoted, therefore I'm going to ignore it." Now, I'm being slightly flippant but not very.

Harry Kemsley: Yeah, yeah. So I do remember, just change the perspective just slightly here, go back to that conversation I had with Di some time ago now when we talked about deepfakes and videos. What I remember from that whole conversation with Di was effectively the human eye is now struggling to really discern the difference between real and fake in the video realm. And she talked about the fact that you can only really find fakes by using other AI. So it's counter AI looking for fake AI- generated deepfake videos. Well, if my experience, I'm not going to keep talking about the scam too much, but with my experience of the scam to go by, that's now not just about videos, that's about all kinds of things. If you send out that email to enough people, as they clearly did, you are going to meet somebody like me who was looking for an email at that moment from that company for that purpose. And that's what worries me is it's very difficult to build tradecraft against something that fits your picture, fits your expectations. Other than, to your point, you should always be questioning yourself. I particularly like the way Amil not only used the line you described in terms of it feels good, check it. He also talked about the fact that when you are looking at a list of news articles and a scroll articles, you could read the one that you pick because of the way it's written and titled, whatever, about is telling you something about the way you are built and your thinking of the world.

Sean Corbett: Absolutely, absolutely.

Harry Kemsley: And I've now watched myself since that conversation quite carefully and I forced myself to read things that I otherwise might not have done. And interestingly, my newsfeed is changing as my echo member is now changing its focus area.

Sean Corbett: Right, absolutely. And going back to the society, so A, we actually do read stuff and spend some time doing it. And you're right, some of the videos now, I saw a clip of, lots of them, and they were ridiculous because they had Donald Trump being dragged off by the police, although that may not be that unusual, but they had some really, really strange ones. But just in terms of you seeing the video, you couldn't tell that they were fake. Now, for us, we're always going to check that sort of thing, but the TikTok generation as I call them, the kids that literally look at 15- second clips, if it's more than 15 seconds they swipe on, they're just going to see that very, very quick snapshot and that's their reality. " Oh, yeah, such and such, this happened." That worries me because that is skewing the realities of life.

Harry Kemsley: Yeah, and indeed as we come on to later perhaps, the democracy we live in. I do remember the person I know who makes movies telling me that the average scene length between different camera angle shots for a particular part of a movie used to be measured in minutes, small numbers of minutes. Today, it's measured in small numbers of seconds. If you actually watch a movie from back in the day, black and white, an average scene would last certain number of minutes. And in that scene they'd had two or three camera angles, period. Now in the same scene they'd have 20, 30, 40 different camera angles because each then clip is only worth two or three seconds, and they believe that's conditioning the human brain to want to get onto the next scene. If you watch an old movie, it does feel a bit comical because of it. I digress. Let's pivot, Sean, from this acknowledgement of the, I'm going to use the word again, pervasiveness of mis and disinformation. We've talked about the reality of that in terms of couple of examples. Let's pivot now into the, so why this important to the analyst? What does this mean for a person who's interested in understanding" ground truth" from open sources and how do they counter it? What are the kind of things that you think of?

Sean Corbett: Okay, so let's take it up a level then to the strategic world that we live in. The primary task of the analyst is to provide the best possible intelligence- assessed information to senior decision makers, or any decision makers actually, for them to achieve their job. So, at the case of the national level, what is China thinking? What do they intend to do? Now, at the time when you are being fed disinformation at the state level, how do you work it out? And the great example at the moment I think, or the best example right now is China. Within the political classes you've got two ends of the spectrum as yours do. But with China, one who says, " No, it's all rubbish. They're a good training partner, we need to make money, we need to get close to them." And the other side saying that they're an existential threat to our Western way of life. Now, which is true? And they're presumably getting fed the same intelligence that as long as it's not politicized, they should be looking at from an objective perspective. So that's part of the: make sure that you provide balanced objective assessments that cover all eventualities, with the understanding yourself as a politician and your skewed bias views. And we're seeing that polarization of views, it's always been there but it's never been so pervasive but also so toxic as we're seeing now.

Harry Kemsley: Is it the responsibility of the analyst to identify that the people they're talking to in terms of the decision makers aren't registering both sides of the argument as they're trying to be presented? Is that the responsibility of the recipient or both?

Sean Corbett: So, having been on the receiving end of that, that's a really difficult one because the analyst needs to have a job, and they need to basically be listened to and be trusted. And you've only got really two or three attempts before you're never allowed again. But you've got to be objective, and you've got to give all sides of the story. Whether your customer chooses to take that or not is not really down to you. Now if you are in a face- to- face situation where you give both sides of the argument, you will very quickly probably, depending on what the situation is, see the direction in which your decision maker tries to take you with just the way they ask the questions at that stage. That's where the personal moral contract comes in and you say, " Actually, sir," or, " Actually, Minister, I don't think that because of this." And the key thing is because of this you've got to qualify what you say. But ultimately people will listen to what they want to listen to.

Harry Kemsley: Indeed. If we just go back into that, though, just one more time. If I were to receive a question or a task that was asking me to go away and do some analysis to come back with some answers, presumably, well, from my recollection, you would then break that down into multiple sub- tasks that would then become eventually the substance of the answer you might create. So that breaking the question down into relevant investigative tasks is the first step, isn't it? Of ensuring you've got a balance of perspectives, balance of sources, a balance of views.

Sean Corbett: Yeah, absolutely right, yeah. And you've got to audit trail them. You have to. Because if you get to a point in your assessment thinking, " Right, I'm pretty sure of this," and you can't work back to, " Where did I get that from? So how have I weighted my sources? What are the sources again? How are they cross- referenced, et cetera?" Then it's your word against somebody else's. So sorry, this is why the IC is so heavily dependent and it's the way that the commercial sector has to go with giving an auditable way of showing their references. How did you come up with that? inaudible-

Harry Kemsley: Yeah, I must confess I've heard that question more than I can remember in the 10 years I've been here at Janes. In the last few years, I've had the question of, " Could we have access to your sources as well as the steps through the analytical process before you produce the outcome?" I don't remember that being the case when I first arrived. So there's definitely been a change in terms of expectation. The other thing I would say, though, there is a moment isn't there between, if you have the time, between when you've been collecting and validating sources, and that's from your breakdown of the questions to the sub- parts of your question, there is a moment, isn't there? Before you produce the report where you come up with your hypotheses, the most likely course of action. They're the most likely answers that you now trying to validate. Isn't that the time when you would probably want to engage with your decision maker on your question?

Sean Corbett: That's another how long is a piece of string question, because it depends on the nature of the question. It depends on what access you've got to that individual, how much of an intelligent customer they are. I don't mean in terms of what's their IQ, but I mean is do they understand what makes up intelligence and the fact that there are different levels of confidence, all the rest of it? So, ideally you'll always do that, you check understanding, and there's the timeliness issue as well. So, in the military domain slightly more than the civilian domain in terms of the agencies, you tend to have to do something quickly. So it's your best assessment in a given time. But again, you've got to give the, " Right, this is my level of confidence and this is why it's the level of confidence. Even that in itself isn't always easy." Certainly there's a couple of great leaders of mine that have said, " Right, Sean, yeah, I get it." Give me your best assessment and you've got to do it. And a good leader will understand that, A, that they put you under pressure and, B, it might be wrong.

Harry Kemsley: And that of course is built on a period of time over which you built up trust and a level of mutual understanding, which of course in my scenario, as I was trying to paint, might or might not be possible in terms of time available. Let me then just twist that just a little bit further into the traditional intelligence gathering methods, which are now much more enabled by technology. You can get sorting engines that do all kinds of work for you that would have taken a lifetime for human being, they do it very quickly. You can get summarization tools that will give you the ability to synthesize that down into bite- sized chunks very fast. And indeed, with certain technologies you can start to detect patterns and use those patterns as a source of what the normal looks like and therefore when something stands out from that, you can spot it more carefully. This is all kind of things that technology is doing for the analysts. That should be, from our previous conversations being about the enhancement of traditional techniques, you are increasing the number of sources you can go to. You are increasing the fidelity of the way you summarize and understand what's in those sources. This is all just about enabling and enhancing the analytical process. When we listened to Amil and previously to Di in this particular inaudible, I got the sense that there were some very specific new tools coming out or we're available that I don't know yet that the analysts that we are talking about, particularly the military environment are engaged with or using. And I'm wondering whether the deficit of some of these more specialist detection tools for mis and disinformation are an area that are a weakness for the analyst we're talking about.

Sean Corbett: That's an interesting question, which I do not know the answer because not being current, but I would be alarmed if they were not all over this. I really would because they've got to be, but it does beg more questions as always than it answers, because who is assuring that extra source of information? So if you've got algorithms that are scraping lots more intelligence to say those algorithms are right? And I've been thinking about this a lot, so you and I have both been playing quite extensively actually with ChatGPT as an AI model.

Harry Kemsley: And others, yeah.

Sean Corbett: And in some areas it's really good. Or is it? How do we know that? But we've learned to trust it because it actually it does seem to bear fruit. But I do always cross- refer. But when you get into something quite specialist like we've got our various different interests out there, it's less granular and certainly less accurate. So, because it's specialist stuff, everybody knows I'm a fisherman, so I'll be asking about, " Okay, what's the relationship between, what temperatures inaudible," all that sort of stuff and you go, "That's not right." So how do we know that the stuff that we trust and know is right? Now, let's not open that particular Pandora's box, but AI has to be part of the solution, but it has to be a part of the solution that we are aware that if you skew the algorithms to start with, you're going to get the wrong answer. Now again, my antidote to that as I've said previously is that just like a human analyst, you rely on their, they'll tell you once if they get it wrong, once they get it wrong twice, not good, three times in year out. So it's the same with the algorithms if it's subsequently proved that... And there has been a few recently quite funny examples actually where you need to put glue on your pizza and things like that because it's just got something badly wrong. So, those will get ironed out eventually and I'd like to think that they will become assured and trusted, but it's a leap of, almost is a leap of faith, the intelligence community are struggling with this right now, an acceptance that we've got to use all sources and we can't do it manually. So much information out there, but how do we trust that AI to actually gather that? And then there's the ethics element which we've covered on inaudible-

Harry Kemsley: I do remember that conversation, I think it was with Keith, Keith Dear, about it's almost negligent now. You could almost be accused of being negligent if you are not using these techniques-

Sean Corbett: Absolutely.

Harry Kemsley: ... and capabilities. But I think it's fair to say, isn't it, though,that given the competence, the abilities of these state actors and disinformation and the spread of misinformation through platforms which we can no longer control, they are" out of our control", not necessarily out of control in the sense of being completely wild, but they're not in our control anyway, that the analyst has to have techniques to deal with that.

Sean Corbett: Absolutely.

Harry Kemsley: And they do also need to start quickly grappling with the tools. But what are the kind techniques, Sean, that we need to rest on?

Sean Corbett: And we're getting to the punchline now, so my favorite word, and people go, "What on Earth are you talking about? You're always banging on about tradecraft." Well, this is absolutely where tradecraft inaudible. So, you can have the best tools in the world that will collect the data and sort it for you and all the rest of it, but then the analyst has to make some sense out of it. And there are, much as I talked about the art of intelligence, there's also the science of intelligence and there's a lot of very well- tried and tested methodologies out there and techniques that have been honed over many years. I mean there's books on it, not that anybody reads books anymore, but there is. I mean, just a few techniques just off the top of my head. Reframing. So that's where you investigate all possible hypotheses no matter how unlikely. You might have seen the two poor astronauts that are stranded right now up in space, well, they had some very unusual noises coming out of their microphone that they sent them back down. Nobody knows what they are. So, almost certainly it's feedback or inaudible it could be aliens, though. Well, it could be, but exaggerating from fact, but you've got to look at all hypotheses no matter how unlikely. And then you can-

Harry Kemsley: inaudible-

Sean Corbett: Absolutely.

Harry Kemsley: ...time available.

Sean Corbett: Exactly. Again, yeah. Then you've got forecasting, which is a tool that uses historical data and past patterns, which I think most of us probably use now to predict future outcomes. " Okay, when we saw these particular situations in the past, this is what happened." And this is what we use at Red Team. We will come up with six or seven different things, say, " If that happens and that happens and that happens, then it's likely that this is the capability plus intent, this is what they're going to try."

Harry Kemsley: Is that almost like a pattern of life analysis?

Sean Corbett: Yeah, that's part of it.

Harry Kemsley: There's a pattern analysis inaudible-

Sean Corbett: Absolutely, so that's forecasting. And then backcasting is slightly different, is you evolve backwards from possible future outcomes, so these are the courses of action, as you and I understand, this is what we think could have happened. Then you work your way back saying, " Okay, what's the likelihood based on what we're seeing?" Which is a nuance, but it's quite different from forecasting. So there are only three, but you can do SWOT analysis and all the rest of it, and the combination of all the above. If you are a professional, competent, and responsible analyst, you'll be using some of those techniques, not all them all the time for the time thing, but also... And again, back to my other favorite subject of all these talking heads on the television who are" experts", that have basically read the Telegraph based on their background go, " Yeah, that sounds about right." They've done none of that work. So, how many times have we heard that Ukraine are about to capitulate? Russia are about to give in and Putin is going to... And all things in that are just hope. And as I always say, hope is not a valid course of action. Now and they're not all the same, of course, because some of them do that work. But as an analyst, you've got to do that work but also show your workings out.

Harry Kemsley: So reframing, forecasting, and backcasting.

Sean Corbett: It's just three.

Harry Kemsley: Three, yeah. Two or three and there will be others. Now, presumably those kinds of techniques, as we were saying earlier about the use of technology to amplify and to enable, we could also be using tools to make our reframing, our forecasting, and our backcasting ever more effective as well?

Sean Corbett: And there will be some apps and tools out there to do it. Again, I'm probably not the person to talk about because I still like Excel spreadsheets, and I still like the estimate process, which even my military friends will be, " Really?" But it works. But as long as you've got a technique that you're comfortable with that you know you're looking at all the different factors that you need to. So do you remember the evaluation of factors?

Harry Kemsley: Yeah, of course.

Sean Corbett: And you'd list them to death and then you'd say, " Right, is that relevant? Is that relevant? Is that relevant?" So, as long as you've got a methodology that encapsulates it all, then you're good to go. But I am sure the clever analysts outside there would go, " Oh, yeah, I do that, but here's the particular tool on the app that I use to do it."

Harry Kemsley: I'm just going to change slightly now, isn't there a value in seeing what disinformation and organization the country is pushing out? Doesn't that give you insights in some regards?

Sean Corbett: Very much so. And this is the other side of the equation. I've been doing quite a lot of work on North Korea actually, and we did a really good podcast with Cristina Varriale. And she was saying as much as it's a closed society and there's not a lot of open source intelligence, and we're going to follow this up by the way, you can tell what to an extent what the nation is thinking by what it puts out there as propaganda. Or if you're in the more nefarious domain, what are they attacking in terms of the cyber perspective? So, if they're probing your electricity system, for example, I'm using that as hypothetical, then you'll think, " Well, okay, they're just testing where the weaknesses are, et cetera." I mean, you could even take it as far as taking it to, and this is an interesting one, but it's more of a physical thing. So, when Iran retaliated for the last strike that killed one of their representatives in Syria, they sent a whole load of drones, all the rest of it, and it didn't get through. Now, you could say the question, " Was that because they just weren't very good or the defenses of the Israelis and allies was good? Or were they just probing and testing?" Don't know. So that's the physical example. But in the electronic domain that is just as reasonable. But from a diplomatic and a propaganda perspective. And will you recall, we had a discussion with Claire Chu who actually went on a trip there.

Harry Kemsley: That's right.

Sean Corbett: And she was saying it was fascinating what people said and how they said it to understand not just what they were trying to message but how they thought as well. So, in the absence of the real intelligence, it does give you a really good idea about where people are coming from.

Harry Kemsley: So don't dismiss disinformation.

Sean Corbett: No, absolutely not.

Harry Kemsley: Use it as a resource as much as any other.

Sean Corbett: Absolutely.

Harry Kemsley: All right, well look, as ever Sean, we've chewed up through 40- odd minutes almost without a blink I think there. What I would want to do with this conversation, though, from where I'm sitting Sean, is I think we need to recognize that mis and disinformation is now the genie out of the bottle. It's happening. And disinformation's been there in one form or by one name or other for a long, long time. But misinformation, which is affecting at the society level is something that we all need to start thinking about. And for me, whether you call it data literacy, or you're just talking about your spidey sense and thinking, " This doesn't feel right," I don't mind which it is, but we do need to start thinking about how we embrace and deal, mitigate with this whole problem because it's not something that's going to go away. It seems to me in the recent years that it's accelerated. We've got more and more misinformation channels and more and more acceptance of misinformation that is driving people further and further away. And I've heard the phrase, " The US is very polarized." Well, they're not unique in that.

Sean Corbett: No, they're not.

Harry Kemsley: And that polarization by whatever means it's being created is being, it seems to me reinforced. I read a fascinating book in recent times, a book, I can't remember the author's name now, it doesn't matter, but it does matter, but I can't remember the name. Everybody Lies, and he talks about the fact that people who you might expect to be looking and reading certain things are actually doing exactly that. They're also on other sites. You've got the same people who are very right- wing talking to left- wing outlets, and vice versa, which does counter the echo chamber argument a little bit, but I'll push that to one side. But what I do notice about that is even if the exposure to other forms of thought, other viewpoints is fleeting, it's better that you're getting some side of its alternative perspectives than none at all. Because the more you become steeped in your own view of the world, the more the marketing algorithms of the world you're living in, the social and other cyber means, is going to reinforce that.

Sean Corbett: Absolutely.

Harry Kemsley: You've really got to get on top of it.

Sean Corbett: And you've got to work at it as well. I deliberately look at some sites that I don't really like just because that's not my politics to see the counter perspective, because it's really easy to get into that filter bubble, echo chamber, call it what you will and just take, " Yeah, yeah, that's what I think." And it makes you feel good. You've got to look at the other side.

Harry Kemsley: Isn't it fascinating though, if you do that how you can feel the adrenaline start to rise?

Sean Corbett: Oh, yes. Oh, yes.

Harry Kemsley: And you start to get more and more agitated by what you're reading and the need to counter it. That is a symptom for me, which is my takeaway from this podcast, by the way, that symptom that you feel in yourself in terms of a reaction, something you want to read and something you want to revoke or to rebut, that is telling you something about your view of the world.

Sean Corbett: Indeed.

Harry Kemsley: And if you are not conscious of that and you are letting yourself just unconsciously be driven along, who knows what this conversation goes. So Sean, your takeaway for the audience?

Sean Corbett: So, my takeaway is that I think this is a really serious issue, and I don't normally use the word existential, but this is a potential existential threat to the Western democratic way of life. We talk about gray zone warfare, which is undoubtedly going on at the moment from the Chinese, from the Russians, from the North Koreans, from the Iranians against Western society, trying to increase that polarization. So this is disinformation, not mis, because it's absolute, and it's working, it really is. You look on any given subject and there are very, very defined viewpoints to the extent that we have had a riots and demonstrations on both sides, which not necessarily always gets... So, you've got the mainstream media have a responsibility here that they're really shirking right now just for clicks. But this, without firing a shot, China, Russia, et al could actually win the next world war just by continuing that, making that happen and getting such a disgruntled population that, A, you don't believe anything or, B, you only believe what you hear because you've heard so much misinformation, disinformation. And there are politicians who are responsible for this as well. Fake news, based on what? How do you know? Some of it might be, but some of it isn't. So straight away you go, " I support that person because they support my, and they say that the others are fake news." So not everybody is as obsessed with getting the right information as we are. So you have a look at, you click on whatever your social media go, " Yeah, right," and then move on. Most people don't care, hopefully, enough to act on any of this. But I could see a position in the next five, 10 years where unless we do something that is... It's got to be education, you can't force it because it just reinforces that polarization that we are in a position where, if nothing else, Western liberal democracies are catatonic because they can't act on anything which we might be seeing now. Or at worse, the Chinese or Russians roll their tanks in and nobody even knows it's happening or can do anything about it. Okay, exaggerating.

Harry Kemsley: Well, we just inaudible a conspiracy theory.

Sean Corbett: Indeed, indeed. But no, being serious for a moment, it is a huge threat to our way of living and to how we run our democracy. And I guess, sorry, this is a long one, but I guess the bottom line is this is why I'm so committed to analysts getting it right and using that tradecraft, because somebody at some stage has to be the grown up in the room and come up with objective, well- balanced, well- researched, fully with all the data assessments and say, " This is the truth."

Harry Kemsley: And that's all about, for me, the open source environment, the one that we are talking about a lot on this podcast, but also that increasingly now a lot of people are becoming aware of the power of that. For the audience listening, for the listener, if you have any questions about this or indeed any comments we've noticed, Sean and I, in the recent months, how many more questions we're getting these days than usual. So, if that's a sign that you want to get more engaged in what we talk about, or indeed make a comment about what we've said, feel free. We are very much up for that. We've had one listener on as a guest, and she was very, very good as a guest, and gave us some interesting insights into the toolsets she's working on and so on. Very happy to engage on that basis as well. So please do send us your thoughts, send us your questions, and we'll happily pick it up. And until then, thank you for listening. Goodbye.

Speaker 1: Thanks for joining us this week on The World of Intelligence. Make sure to visit our website, Janes. com/ podcast, where you can subscribe to the show on Apple Podcasts, Spotify, or Google Podcasts, so you'll never miss an episode.