Joseph Hatfield PhD and David Gioe PhD join Harry and Sean to discuss the challenge of defining open-source intelligence alongside other intelligence disciplines and why now might be the right time to redefine OSINT.
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. Before we start this podcast episode, just a quick explanation that we're going to split it into two parts. The first part we'll play now and then we'll invite you back to join us for the second part very shortly. Hello and welcome to this episode of World of Intelligence by Janes with your host Harry Kemsley, and I'm glad to say back with me is Sean Corbett, my co- host. Hello, Sean.
Sean Corbett: Hi, Harry. Good to be back.
Harry Kemsley: Sean, you did me a great favor recently, put a front me a paper that had an interesting title. There's no such thing as OSINT, and I thought that would be a great place for us to have a conversation about OSINT. In addition, I had a request from a listener about could we explain what OSINT was about? A more inaudible question. These two things have come together. Absolutely delighted to therefore welcome to the episode the first guest, Dr. Joe Hatfield, the assistant professor at the United States Naval Academy Department of Cyber Science. Hello, Dr. Joe.
Dr. Joe Hatfield: Hi. Great to be here. Thank you.
Harry Kemsley: Yeah. Thank you for the article, by the way, of which we're going to have more in a moment. In addition, our second guest today is Dr. David Gioe, the British Academy global professor and the visiting professor of intelligence and international security, Department of War Studies at King's College, London. Hello, David.
Dr. David Gioe: Hi. Thanks for the invitation.
Harry Kemsley: Thank you both very much indeed. Right. Given that this podcast is principally about the power, the potential of open source intelligence or intelligence derived from open sources, and that we want the audience to understand that and how we might engage with it, this topic is of great interest to me. I'm going to get Sean in just a second just to remind the audience how we've previously defined what we're calling open source intelligence. Then Joe, I'm going to come to you first in terms of your perspective on that, particularly in the paper that you've written. Sean, starting with you, can you, for the benefit of the listener, remind us all what we consider the definition of OSINT to be?
Sean Corbett: Absolutely. It's worth noting, and I think we'll get into this straight away, that definitions may vary, which I think is partly the thesis inaudible-
Harry Kemsley: Is that a caveat up front, Sean, just for the inaudible-
Sean Corbett: Absolutely. I'm looking forward to this one because we tend to agree furiously on these, but I do like a good argument, as you know, and I'm hoping there's something between us there, but I'm not sure there will. But anyway.
Harry Kemsley: Let's see.
Sean Corbett: Those of the regular listeners from the early days, we characterized open source intelligence in four ways. Firstly is the inaudible obvious that it's got to be derived from publicly or commercially available information. That can be such a broad spectrum of stuff, but it has to be for us acquired legally, and certainly for James and myself, ethically, and we've done podcasts on the ethics before. The legal element is important. Then the third and fourth really elements definition applies for me to all intelligence. Firstly, it has to be applied to a specific problem set or question. The third thing is it has to actually add value. Reporting is slightly different to me from what intelligence is. That in a nutshell is the four characters of it.
Harry Kemsley: Thank you very much, Sean. That's a reminder for our listeners who may have listened to some of our early podcasts, the definition that we've been working with for the last period of time. Joe, over to you. What's your view on the notion of open source intelligence, particularly in light of the paper that you wrote?
Dr. Joe Hatfield: Thanks, Harry and Sean. Well, there is no such thing as open source intelligence. That is my thesis. I do not think that it is a coherent concept. One thing that I have to say right off, although that title is provocative, it often is provocative, because people assume incorrectly that when someone states that there is no such thing as open source intelligence, what they mean is that it has to have secrecy involved. In order for it to be considered intelligence, it has to be in some sense classified. I'm not making that claim. I'm not saying that publicly available information is not useful for intelligence, but what I am claiming is that the very notion of OSINT or open source intelligence that we've been using for decades now, it actually traces all the way back to the 1960s. In my paper I do give a history of the concept of OSINT, its frequency of use, when did it arise, et cetera, et cetera. If you're interested in the paper, it was published this year in the International Journal of Intelligence and Counterintelligence. But that's not my claim. My claim is that the idea of OSINT is fundamentally incoherent. I do it in a couple of ways, but one way in which I do it is I basically take a look at how we're defining OSINT relative to how we define every other int. What you find is that in the system of categorization, the taxonomy, if you will, of the various ints, what happens is you've got a very clear rule that derives or generates the taxonomy for all the other ints. Then if you grab OSINT and add it to it, the same rule cannot be applied. You have to actually introduce a new rule. For example, IMINT or SIGINT. These are human. These denote the source of the information. Is it a human source? Is it from an image, et cetera, et cetera. Whereas for open source information, supposedly it's sources that are openly available. Well, that could be a human, it could be an image, it could be any of these things. But now we're adding a new category that violates the taxonomical rule that we use to derive the first system of classification. Now, that in and of itself is not necessarily an objection, because someone could say, " Okay. Fine. We have a more complicated taxonomy, but it's worth it to add this concept." But if you actually push on that new rule taxonomy, it turns on some concepts that are very, very precarious. For example, it's typically defined as information as openly available or it's defined as publicly available or in the public domain, so it needs to be in the public domain. The first thing that we realize is that it's being defined negatively. In other words, it's being defined by that which it is not, so it is not classified. But whenever as a philosopher, I did a bachelor's degree in philosophy. I love philosophy. When you find yourself defining things negatively, then you are in a little bit of conceptual hot water, because you're unable to say what it is. You're just able to define it as kind of a leftover concept. I call it in the paper a junk drawer concept. We all have this junk drawer in our houses that is full of rubber bands and paper clips and a hammer and whatever else. The reason why the hammer is not in the cutlery drawer or the silverware drawer or whatever is simply because it doesn't fit in that categorization. What correctly goes in the junk drawer? Well, anything that doesn't fit everywhere else, and effectively that's what OSINT is. It's a junk drawer. When you find yourself defining things that way, that's problematic. Then another point I would make is that even the very notion that it's publicly available kind of papers over or covers over or hides all kinds of problems with that. For example, if I don't speak the language, it's not really available to me without some sort of translation service. If I don't have the requisite background knowledge to actually make sense of it. For example, I'm not a physicist. I don't have much knowledge in physics, so that academic paper might be publicly available, but I have no means, I have no way to actually interpret that in any way whatsoever. Is it really available? It's not a very robust sense of availability. There's also things like if it's difficult to find. I might be able to interpret it, but it's on the dark web. Well, it's publicly available, but I need to understand onion routing. I need to know what the Tor network is all about, et cetera. In what sense is it truly available? Effectively, my critique from the conceptual level, and then I have a historical critique, which maybe we can go into later. But from a conceptual level, it's a junk drawer concept. It introduces an additional rule to the taxonomy. It is not in my opinion necessary. Later in the paper I argue that human source information is HUMINT, whether or not it's publicly available. Imagery information is IMINT, whether or not it's publicly available. Effectively, what I'm trying to say is we only need the original taxonomical rule. We don't need to introduce this new thing. This new rule is full of problems when you push on it, and effectively that's kind of the first stage of my critique. Maybe I'll stop there.
Harry Kemsley: Joe, it's fantastic, because you've lined up for me at least 300 things that I want to tuck into, so we're going to spend the next four or five hours. Listeners, buckle up, get yourself a coffee, because this is going to take a while. Really looking forward to this. But before I launch, and I can see Sean is leaning forward, ready to go. David, let me come to you next. Your perspective on this concept of open source intelligence.
Dr. David Gioe: Yeah. Thank you. Well, just to react to Joe quickly, I'll note I was keeping tabs here and he said taxonomy four times. This is what you get when you have the philosophers using words that nobody knows, and actually I appreciated this paper from Joe, because it was the first one I actually understood. That was great, but I'm also desperate to know what's in Joe's junk drawer at home.
Harry Kemsley: So am I. So am I.
Dr. David Gioe: I doubt it's limited to rubber bands and hammers, but it might come out later. Anyway. Well, I'm going to keep my taxonomy count, because Joe is not done yet, I'm sure.
Harry Kemsley: That's two, by the way, David. That's two. You've said two. Joe had four.
Dr. David Gioe: I'm quoting him. I'm quoting him. But for me, I'm just astounded. I think Joe has a very specific argument and I'm kind of just along for the ride, because I'm just astounded at the definitional differences. What's interesting to me is as a scholar of intelligence, I never ever get into the definitions of intelligence, because as soon as I say the definition is this, someone says, " Well, what about this? What about that?" Actually, I just avoid these definitions altogether, because they're so fragile and they mean different things to different people. I have a longer origin story that we can get into at another time, but when started, when I emerged from the secret spaceship in 2011 and I was like, " Wow. There's a whole world of information out here and it's not classified, but it's useful. What do I call it?" People just settled on OSINT, as Joe says. Just a thing that it's not secret, so it must therefore be OSINT, and that's kind of where we got to. But I just wanted to kind of pause and for me, if someone says, " Can you do a paper based on OSINT sources?" I would say, " Sure." But for me, that just means using the internet, just using publicly available or commercially available sources. I was talking with a student who works for an OSINT sort of company and he was talking in terms of tools. He says, " Yeah. I'm conversant in 126 different OSINT tools." To me this was like when C- 3PO tells Han Solo he knows X number of protocol languages. It just doesn't compute. I have no idea. I couldn't name one of those 126 tools that this student was talking about. I think it just goes to show you how far the field, if you like, has come from using non- secret sources to using dedicated tools that have been developed. One just quick anecdote. I got a text from a friend of mine who teaches intelligence analysis at Brunel University. Hey, Kristian Gustafson. If you're listening, this Bud is for you. Christian said, " Hey, just take a picture of wherever you are right now and just send it to me." I said, " Well, that's weird. What are you doing?" He said, " I'm teaching OSINT and my class is going to find you." I was picking my son up from school. I didn't take a picture of the sign. I tried to make it a little bit harder than that.
Harry Kemsley: Little bit more difficult than that.
Dr. David Gioe: But I just took a picture of the building and I was like, " Hey, this is where I am." Then he said, " Great. Thanks." Then the students set about trying to find me using publicly available tools. If you put a gun to my head and said, " Find where you are with this tool," I wouldn't be able to do it. Really there's a whole discipline that emerged around these ideas, but as Joe says, we're not entirely sure what we're talking about and often they're defined in the photo negative of what we really mean.
Harry Kemsley: Okay. Again, lots of things I want to pick up on there and I'm now leaning forward in my straps, Sean, but I'm going to give you the chance to go first. But I may put a time limit on how long you've got to answer your initial reactions to that, Sean, before I get my go. Sean, over to you.
Sean Corbett: Oh, good, good. Because Joe had his hand up there and I probably wouldn't get a word in edge wise. This is fantastic discussion, by the way. There's so much in that. We could spend the next 40 minutes just talking about definitions, actually. While we will not do that, they are matter. I think the natural tension, if you like, is the difference between the conceptual and the practical. You could bring it down actually in terms of how do we do intelligence to a cultural element. I'm looking at if we invented intelligence today, and there's quite a lot of papers out there right now, we certainly wouldn't do it how we did. As you said, Joe, used since'50s or'60s open source intelligence, but it's still not accepted formally from a source perspective as saying we can use stuff that is publicly or commercially available, because it's got to go through this whole process of assurance and all the rest of it. Now, the counter to that is yes, we can do all that, and we must come back and talk to that. But I think the nub of it really is that there is so much data out there now, and you've heard the word democratization of data. We have to be able to leverage it. You might have heard of talking about OSINT as being the intelligence of first resort. The community is starting to talk about that now, which is really interesting, but they're lightyears away of actually achieving that. But to say there's no such thing as OSINT kind of just in terms of culture takes an entire huge raft of sources out from the equation that you can't use if you're within the intelligence community, and that's what we're trying to do. I think what I'm doing in a very convoluted way is I'm semi agreeing with you in terms of there is just intelligence. But in terms of leveraging the full gamut of what's available, we have to term something that you can do by commercially or publicly means. Just one thing I would just start gently push back. On all intelligence collection has its own tools and requirements and you need to be a specialist in some way, and open source intelligence to be a responsible OSINT provider rather than just Googling things or saying what you first see and the disinformation, misinformation we get in there as well requires techniques, tactics, and procedures to make sure that that information is assured as well. I could go on, but I'll stop.
Harry Kemsley: All right. Very good. All right. Before we go to the next round of this, now that the conceptual lines are somewhat drawn, let me see if I can summarize a little bit what I've just heard. By the way, gents, if I get it wrong, please tell me because I'd like to get this right. I think what I heard you say, Joe, is that we've created a set of words. I'm not going to use the T word. A set of words that we've defined as something that is unnecessary. It's almost a redundant statement, that actually I quite like the point you made about an image. Whether I can get it from a closed classified source or a source that's publicly commercial available is still an image. If I derive intelligence value from that image, whether it's from a closed or open source, it is an image, and therefore it's an image process of analysis that's derived intelligence. By the way, I don't like the phrase, or let me go back, I've come to dislike the acronym OSINT, because it's being misused in my opinion. Ask me later why I think that. I prefer that I can derive intelligence value from open sources and then we summarize that into OSINT. David, I absolutely agree with you. There is almost a religion that's been built around the open source environment with tool sets that made people believe that they are now intelligence analysts by using these tools. Actually, that isn't necessarily a bad thing. But to connect that to what Joe was saying, this continuum of available information from which you can derive intelligence with tools through a recognized trade craft is, I think, a universal. It's not a thing that's unique to the open source environment anymore that is a classified. Then Sean, the one thing you didn't say, which I assumed you would or that you'll come to later, is that trade craft, which is about bringing together great judgment, great process, and increasingly technology these days, is about deriving an answer that's as good as you can make it, reducing the uncertainty from whatever sources you have available. But I'm sure, Sean, you'll use the trade craft word at some point during the course of this conversation, because without it it wouldn't be a podcast with Sean involved.
Sean Corbett: Absolutely.
Harry Kemsley: Joe, I summarized your thoughts into essentially that you're describing this continuum that is wherever the source, it is nonetheless an image or it's a signal of some sort that you're going to use and derive intelligence value from. Is that a reasonable summary of what you were saying?
Dr. Joe Hatfield: Yes, it is. I think the central utility of, and I say this in the paper, of this idea of OSINT is precisely what Sean was worried about. That is that we have all this wonderful and ever- growing amount of publicly available information that it would be irresponsible of intelligence analysts not to take into account. If you trace the history of intelligence, it turns out that we've been taking public information into our account for decades. It's not a new thing. But the sheer magnitude and sophistication of what open source intelligence outfits can produced today is on many orders of magnitude. What was once the purview of states is now clearly absolutely available.
Harry Kemsley: Satellite technology is a great example of that. Just the fact I can go to a credit card machine and put it in and get an image. That small, high accuracy over a credit card tells you that that's happened for sure.
Dr. Joe Hatfield: Absolutely. The utility of the word OSINT was that it packaged all of this new stuff in a way that made sense to the guild that is the keepers of the secrets, the people, as Dave says, who were behind the inside the skiff. We call it in another paper of ours the cult of the skiff, because back in the day, the people inside those walls of that secret space really did have better sources. They really did have superior knowledge, because they had openly available information and very, very exquisite human sources. They had signal sites and things like that that just other people didn't have. The utility of this idea of OSINT, I think there is no such thing as that, but that doesn't mean that the use of the concept, the invention of the concept, it doesn't mean it did not perform a very valuable function. What I'm arguing is that that utility is now, it's one. There's no responsible analyst today that would not take into account publicly available information, but now that concept has become a hindrance, because really it was kind of a fake concept. It was philosophically not very well understood and derived. It did perform a useful function historically, and actually the way in which it was packaged allowed it to do a number of things to people within the guild. It was a great way to do it. I think it did four things. One was it neutered a potential threatening rivalrous development, which is all this open source information that could rival. By classifying as OSINT, it sounds like that's just another int and we can be experts. We're the experts of the ints and we can have OSINT analysts, even though it makes no real sense to say I'm an OSINT analyst. How many bazillion different techniques would you need, tools would you need, areas of inquiry? It just simply makes no sense to say that, but yet we have on CIA's website, we have hire an OSINT analyst. USA. gov. You can hire an OSINT analyst as if that is semantically meaningful at all. But by doing this, it neutered the potential threatening rivalrous development. The second thing it did, I think, is that it expanded the jurisdiction and remit of intelligence organizations to this broad, vast amount of data. As new things were coming out, they could categorize it as OSINT, and therefore they could ask for more resources, funding, et cetera in order to tackle it. T. He third thing would be that it actually does, in a weird way, it preserves secrecy fetishism. It preserves secrecy fetishism, and the reason why it does that is because... Think about how analysts are trained. Let's go look at some IMINT, like an all source analyst. Let's look at some HUMINT. Now we've got our OSINT, as if proportionally it fit just another int category. But if you actually look at the amount of information that's publicly available today, it dwarfs anything going on inside the skiff by many orders of magnitude. But if you lay them out like the puzzle pieces all weight about the same and you need to take into account, you got to go check your OSINT as if that was just an equal category as going into a human database or going into imagery or whatever. It preserves secrecy fetishism, because it appears as if the secrets really still matter that much, when in reality, especially the strategic level, they kind of don't matter that much. Then the final thing it did is what I'm saying now. It's this notion of OSINT successfully packaged all this new information to an audience that allowed them to understand that this is valuable. I don't want to take away from that fourth bit. I think that fourth piece of directing our attention to the wealth of information that's publicly available, that is absolutely valid. But it does not mean that we still today need to remain saddled to this incoherent concept, this junk drawer concept. I think we ought to get rid of it. It's served its discursive purpose. It has served it. We've climbed the ladder now. As Wittgenstein said, we throw it away behind us and now we're up on top and we know how to do these things, so we don't need it anymore. Let's de- homogenize OSINT, bring it back to its original taxonomical sourcing. The HUMINT source that the BBC has in Tunisia prior to the Arab spring, who can tell them about what's going on in the marketplace and who doesn't want to be on record and wants to be deep background, and so they use clandestine means to meet. That's a human source. That's a human source and it ought to be called a human source. That's effectively what I'm arguing.
Harry Kemsley: David, again, I'm going to summarize a great and very eloquent piece, Joe. That sounds like what we're saying then is that the concept of OSINT served a purpose, but that purpose is now redundant, that therefore there is no requirement for us to now differentiate any form of int from open or closed sources. What's your thoughts, David?
Dr. David Gioe: Yeah. I think as Joe says, the int on the end of it is a little bit distracting and I feel like you've got IMINT, or the cool kids, by the way, are calling it GEOINT now. The IMINTers get very upset when go back to U- 2s and spy planes and that sort of thing, but I'm old school. I still call it the GRU and I still call it IMINT. Anyway, but you've got your IMINT and you've got your SIGINT and you've got your HUMINT, and those are all recognizable members of the team. They're different bits. Then you put OSINT in there and people are like, " Oh, boy. Is that one of the cool kids?" I don't know, it's lesser. Maybe because it actually comes out of something that wasn't super secret. Just as an intelligence historian, we all know that human sources need to be kept super secret. We know how much money it costs to deploy a satellite or to develop skunk works and spy planes, and we got to keep all that hush- hush. But where does OSINT come from? I would guess that some of it comes from the Foreign Broadcast Information Service and just news monitoring. I love the Central Intelligence Agency, but I do have to say that one of the great cons is when they say that they've released millions of records, what they've actually released is press clippings that they've used. If you go to the US National Archives and you go to what's called CREST, which was the CIA records and whatever review tool, and you would ask it to search. It would come up with Economist articles or Washington Post articles, and they would say, " We declassified this." No, no, no, no, no, no. This was never classified in the first place, so let's not give yourselves credit for that. But anyway, I guess just the point that I wanted to make was more about how the government would harness OSINT. One of the real problems, I think, is that we now want to have another separate agency, because the SIGINT people have their own agency and the HUMINT people have their own agency, and so now the OSINT people must have their own agency. I'm very concerned about this, because I don't want to have some super SIGINT nerd and some super HUMINT nerd. Why do we have all source analysts who are at one agency, be it the JIC or CIA or wherever, and then have the OSINT people at their own separate thing looking only at non- classified sources? If I were king, I would teach analysts to use unclassified sources as equally useful as classified sources, but just making front and center that of course, the classified bits need protecting. That's always the trick. If you homogenize it too much, then you think, " Oh, well, I can share this piece of unclassified stuff, so maybe I can share this piece of secret stuff." That's where you're going to run into problems.
Harry Kemsley: Very good. Sean, I can see you're dying to jump in. I'm going to let you do that, but I'll keep you short, because I want to take it in a slightly different direction, which has come to me from a recollection I've had from reading the biography. You can't see it for the listener, because they haven't got a video. This is the book of the history of Fred T Jane, the originator of Janes back in the late 19th century. I want to share with you an insight from this book that I think might be illuminating to this conversation. Sean, I'm going to give you release for a minute or two, but no more. Over to you.
Sean Corbett: Okay. Very quickly, because we have already pivoted to the what does this mean, which is really important and I would like to discuss that. But I do think there's something in here about delineating in their own mind the difference between intelligence collection and analysis, because all the different agencies are basic collections agents who are very, very good at what they do and very specialist. But they tend to pass it. I know some of them do their own analysis, but is it single source or is it multi- source analysis discussed? What that really leads to is that do we bucket OSINT as a collection in terms of what source it is or in terms of the output? I guess the way I look at that is DIA, they call it all source. It's not. It's multi- source. That's another discussion for another day. But they use sources from everywhere, whether that's CIA classified or unclassified, which they struggle to use because of all the policies and all the rest of it. But let's park that anyway and get onto how do we organize how to use open sources in terms of the intelligence community, because there are different models. You either can say, " Okay. Within each of the analytical agencies you can have an element that uses open source," and to an extent, some of them inaudible do. NGA uses commercial imagery. Of course, it does. CIA, I'm assuming, uses reporting that we've talked about a little bit, human reporting, et cetera, et cetera. But if you do that, you're then stove piping it again and say, " Well, I do openly available information collection, but only from commercial satellite or only from human or only from SEMA in commercial signaling equipment." Which then stove pipes it even more, but if you do counter, which is the most ridiculous thing, and I know apologize for anybody who's written on it there, but there are some fairly distinguished people have written, " We need an OSINT agency." It's absolute nonsense, because A, it does exactly the things that Joe was talking about and it almost negates the all source or multi- source intelligence. B, it's impractical anyway, because it would cost lots and lots of money. You'd need to go through all the checks and balances judicially and all the rest of it, and then someone would want to own it. A lot of it is about the ownership. For me, the answer is to regularize OSINT in terms of cohering, and now I am going to talk about the trade craft, the trade craft. Providing that piece of information has been properly curated, verified, validated, and incorporated with all the rest of the information that whatever classified is, and then weighted and the analyst who understands what they're talking about goes, " I think this because of that," and then sourced. Then it should be absolutely inherent in everything we do from the intelligence world. I know you want me to shut up now.
Harry Kemsley: No. That's fine. That's good. I do want to pivot, because as you've been talking, I'm reminded of a line from the story about when Fred T Jane back in the late 19th century was doing what he was doing in the pubs around the port of Portsmouth, a naval port in the south of the UK. He was collecting from open sources, sailors, understandings of the various aspects around ships that had been drawing during the day. One of the things he said in public was, I'm going to quote, that, " The King Edward VII class battleships were essentially prize gooseberries." Meaning all very well to make the neighbors stare, but of limited practical use. He derived that statement deliberately to provoke the Royal Navy into buying his book and getting into his naval war game. Now, that's not the reason I'm bringing that story up. What he did for many, many hours in the pubs and bars around the port was collect from multiple open sources, sailors, a series of insights about the ships' capabilities that were then compared to other battleships that came in from the Japanese, the American, the Russian navies. Then created a book that said, " Operationally this is a gooseberry. We should get rid of this or replace it with something more useful." Now, that was from open sources. Now, Janes doesn't necessarily spend quite as much time in bars plying sailors with beer, at least not officially. But the point here is we are deriving intelligence value from open sources, but I'm quite taken with this concept now of this continuum. Pause. What I'm therefore going to now posit is, so why did the word, the acronym OSINT become part of the taxonomy, the vocabulary, the lexicon of the modern time? When did it start? I don't remember OSINT being discussed when I was in service 12, 15 years ago. It may have been, but it may not have been called that then. I've got a feeling, Sean, that this goes back to something you and I talked about before about the arrival of certain technologies that allowed us to exploit publicly available information in a way that we couldn't. David, you said earlier about the 126 tools your student had access to. Those tools weren't available a long time ago, because the technology didn't open the box. I'm going to come to a point shortly, I promise. But I'm also reminded of the conversation we had with that senior officer who had worked in diplomatic circles who said, " What's all the fuss? We only operate from open sources. We've only ever operated from open sources, which is the diplomatic channels of human interactions. Why are we making it into some religion?" I wonder, therefore, Joe, back to your four reasons why this has become useful, but now redundant, whether actually the genesis of this was the technology that enabled us to grapple with the velocity, the variety, et cetera of open sources in a way that we couldn't before we started talking about OSINT. The minute the technology allowed us to do it, it needed to have a science, it needed to have a religion behind it. Does that seem fair? Is it really just a matter of technology?
Dr. Joe Hatfield: Well, that's a very good question, Harry. In my research, and I have two charts in the paper that actually statistically show my research into the origin of this term. What I would say is you're partially right based on the historical analysis I was able to uncover, but it's a little more nuanced. The term I traced all the way back to the 1960s. I found US Congressional reports in the 1960s that mentioned the term open source intelligence as it relates to some domestic things that were going on. The term goes back decades before the worldwide web and the kind of explosion that we saw in the 1990s, but that doesn't mean that it was in everyday use in the lexicon of regular people as well as just analysts and so on. In fact, what I discovered was that it was precisely, and this gets to your point, which I think you nailed it. It was precisely in the 1990s that you see an exponential rise in the use of the term OSINT. Early on, it wasn't a uniform spelling. Sometimes it was capital O, capital S, and then lowercase int and things like that. Sometimes there was a C in it. But basically, it was around the 1990s that you see it take off and it just grows at a very rapid, almost exponential curve. You would think, " Okay. That's in general." If you look at something like an Ngram from Google, which can trace published information about actually all publications that are indexed by Google, you'll see exactly that path. The other thing I did was I went back and looked at something that was a little more recent, but prior to the 1990s, which is I took the two top peer reviewed intelligence journals, so Intelligence and National Security being one and the International Journal of Intelligence and Counterintelligence, excuse me, being the other. What I did is I looked through the entire history of their catalogs, both of which started in 1980. Oh, gosh. Let me see here, 1986. They both conveniently, their first issue was in the same year. I have a nice bar chart where I talk about when OSINT was mentioned. These are scholars and practitioners of intelligence who are using the term OSINT or not. What you see even in these journals is similarly, right around the 1990s, you see this explosion of use. You see this massive, it's an expansion. Then what I did was I separated out not just was OSINT mentioned in the article, but does it appear in the title of the article, because if it appears in the title, that means it must be a lot. Most of the article is about OSINT.
Harry Kemsley: Is about it. Right.
Dr. Joe Hatfield: Yeah. It's not just some little footnote somewhere, but it's actually the core. What you see is that it starts appearing in the title around the late'90s, 2000s, and then it gets ever more present within the titles of articles all the way up to my article, which of course, is there is no such thing as open source intelligence. I think you're exactly right that it's not that the origin of the concept began around the'90s when the worldwide web really became publicly available, because it began decades earlier, but it really exploded. Basically, when the environment was right, it bore fruit. The example I give in the paper is it's like one of those insects that bury under the ground for 17 years, and then when the environment is right, they come out en masse and then they take over. It's like that. It's very much like that.
Harry Kemsley: We've effectively, in the last half an hour or so, we've started to equate OSINT with the junk drawer and an insect that buries itself, maybe like a termite before it leaps above the ground and flies. Got it. David, I'm going to come to you now in terms of the academic study of, I'm going to say so- called open source intelligence for the benefit of this conversation. Is it something that we need our students to understand in the context of intelligence as a discipline, or do we need to separate it in some way? Is there something about what we're calling open source intelligence that needs a special study, a separate study to, " Intelligence?"
Dr. David Gioe: Yeah. I'm going to make myself very unpopular now, because students nowadays, especially on a master's program, want skill development. They want to know. They want to come get a master's degree and they want to learn how to use these tools for employability purposes. Of course, what the students want, what they're willing to pay for in tuition shapes the sorts of hires that departments make. Departments are now hiring people who are gee whiz with these tools, and I think that's a weak way to hire someone. But that's in response to what students think they need, because unless they can come out and they say, " I have this skill." I went in not knowing how to use IntelligencePython, and then I graduated knowing how to use IntelligencePython. Therefore, employer, you should hire me." It reminds me of when I was growing up in the 1980s, you turn on the television and there were signs for VHS repair and Betamax repair. What if you were teaching people how to repair a Betamax machine? For me, I would rather teach students how to think, how to critically assess, how to analyze, how to wade through lots of material than a specific application or skill, but I'm losing that battle. The one thing I did want to say, Harry, coming back to when you were in the service. I joined the US intelligence community in 2001, and at that point OSINT was used. Joe will just forgive me for the shorthand, although I think he's agreed, maybe we've just lost the war on this one. OSINT was used in the worst possible way imaginable. It was used for liaison engagement, because you'd have a meeting with some foreign intelligence service and you'd go to the foreign disclosure people and you want to be like, " Hey, look. The sky is blue." They're like, " You can't say the sky is blue, because that comes from a multi- billion dollar satellite, and then this country's going to know that we have the satellite that can tell us that the sky is blue, because it measures the number of electrons in the clouds or whatever it does." People say, " Well, okay. Can't you just find it in OSINT?" OSINT became the substitute, the ersatz classified thing that said what I wanted to say in the secret world. Oh, look, my wife, it was a great story. She got in trouble. Joe remember. She wasn't allowed to say what she wanted to say about smuggling networks at some NATO meeting in Brussels. The foreign disclosure person said, " Well, look, here's an Economist map that basically says the same thing." Well, it said a whole bunch of other stuff, too. It was right. Then my wife got tons of abuse for this map, and she's like, " Well, it's not my stupid map." But the point is it became a parallel unclassified universe that was good enough to get your secret point across with foreigners or with uncleared people. That's like using, I don't know, your Rolex watch to open up cans of paint with the lugs. You can do it. That's an application, but it's the dumbest possible application. I think where we've come now is actually at least showing that this is a powerful approach, methodology. Harness it, call it what you want, but we're now way beyond the days 23 years ago where it was just say it with OSINT, and that way everyone will be happy and the foreign disclosure officer can go back to playing tic- tac- toe on the computer. I just want to point out that OSINT itself has had a utility transformation, but we might overdo it and make it its own separate thing and then pigeonhole it in some other place, which Sean has already mentioned.
Harry Kemsley: Joe, go ahead. Joe.
Dr. Joe Hatfield: Well, one of the things I wanted to point out was that my proposal is, again, not to downplay this information at all. It's just to bring it into the traditional taxonomy. You actually see this all the time. I've been an all- source intelligence analyst. Now is about as good a time as any to also put my disclaimer in that my opinions on this podcast are my own and do not reflect the Department of Defense, the Navy, or the Naval Academy. But anyways, in my role in the DOD as an intelligence officer for 20 years, one of the things that I did, I would do the analysis of for example, human intelligence reports. When I was taught how to analyze and interact with human collectors and all of that, I knew that what I would need to do is I would need to carefully examine the source summary statement provided by the case officer to say this person had firsthand access or second. How reliable were they in their reporting and all that. But because we don't classify OSINT in the same category, none of those human analysis methodology was thought to be applied to the New York Times human source. What you end up with is you end up with people having very sophisticated, what you're calling analytical trade craft. What we all call analytical trade craft, but on the classified side. But we fail to bring those analytical standards over to a lot of our open source. For example, if I had a question about the testimony of a human agent and Dave Gioe was my collector, I could reach out to him and say, " Hey." I could have a dialogue like a question with the collection manager or whoever. But how many reach out to the BBC reporter or reach out to the New York Times reporter and say, " Hey, I know you've got this person on deep background, but I need to know this as an intelligence analyst." I will tell you the answer is zero. I have a very good friend who works for the BBC, and I asked her as well, " How many times has somebody in British intelligence or anybody else reached out to you to check the validity of the source?" She's worked there for a long time. The answer was zero.
Harry Kemsley: Zero.
Dr. Joe Hatfield: It's not just a philosophical argument. There's a practical argument. We're hiring people who are so- called OSINTers. That is incoherent. We're hiring the wrong people. We're not applying analytical trade craft that we should be applying in unclassified areas precisely because we're hanging this albatross of this concept of OSINT around us. We need to free ourselves from this in order to be able to be not only taxonomically complete, but also methodologically sound.
Harry Kemsley: Okay. We'll take just a short pause there. That's the end of part one. Please do join us for part two very soon. Thank you for listening.
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.