Home Page
Information Communication & Society Feature Article - Vol 2, Num 2, Summer 1999

Current Issue
About iCS
Past Issues
Reviews
Contacts
Resources
How to Submit
Subscriptions
 
 
Contents


Information Warfare and the Future of the Spy

Philip H.J. Davies

Department of Sociology, University of Reading


ABSTRACT

This article examines the impact of the new ICTs on the collection of covert intelligence and covert political actions undertaken by national intelligence agencies. It is argued that there exist two distinct doctrines in the literatures of intelligence and information warfare concerning the future relative importance of information from human sources ('agents') and technical methods (signal interception, overhead imagery and the emerging field of clandestine penetration of networked information systems). The arguments in favour of human and technical methods are examined in the context of information warfare techniques and technologies, as are covert action methods such as disinformation, disruptive action and 'cyber-sabotage'. Certain civil liberties implications of ICT-based strong encryption are also examined. The article concludes that what is required is a greater emphasis on integrating
human and technical methods into a unified whole, especially where human methods can provide opportunities which can be further exploited by technical methods.

KEYWORDS

action, covert, espionage, infosec, infowar, intelligence

Is information warfare a completely new form of conflict that exists because of the burgeoning global information infrastructure or is it merely a new dimension of an old form, like spying, whose origins lie in the 'grayware' of the human brain? (RAND Research Review Information Warfare and Cyberspace Security 1995)

There were cowboys ever since there were computers. They built the first computers to crack German ice, right? Codebreakers. So there was ice before computers, you wanna look at it that way.(Gibson 1987)

INTRODUCTION

Information warfare represents one aspect of what has come to be known as the 'revolution in military affairs'. While this somewhat millennial view of how the scope and nature of conflict has changed since the end of the Cold War and the Second Gulf War is more generally accepted in North America than the UK, there can be no doubt that recent information technology, from computers through to satellites represents a new arena of threats and opportunities, and if military affairs must change to keep pace, so must virtually all other aspects of the national security infrastructure. Amongst the elements of national government facing the burgeoning new 'infosphere', intelligence more than almost any other is about information. However, apart from Winn Schwartau's offhand rhetorical arabesque that 'spies are the original information warriors' (1996: 68), the role of secret intelligence in this emerging strategic environment is something which has received too little attention, and what thought has been given the matter requires careful reappraisal and reassessment.

What has emerged in the literature has been two competing views of intelligence. One view stresses the high-tech opportunities of cybernetic penetration and disruption, and is concerned with digital espionage and sabotage by Gibsonesque 'console cowboys' who 'hack into' information systems in a globally networked cyberspace. Since the Second World War the primary sources of intelligence 'take' have been technical methods, or TECHINT, traditionally composed mainly of signals intelligence (SIGINT) and imagery intelligence (IMINT). To these the information warfare enthusiast would add cybernetic operations, what might be called, for lack of a better term (although a worse one is hard to imagine) HACKINT.1 The other approach, however, points to the distributed, easily concealed and pervasively low-intensity conflict world of terrorism, proliferation and transnational organized crime, and anticipates a new age of human intelligence, or HUMINT. Thus intelligence policy-makers, seeking to redesign the world's intelligence communities for the near future and trying to allocate increasingly scarce intelligence resources, are confronted with two competing views of the coming decades, that the future holds either a fast-paced, high-bandwidth TECHINT world, or that it holds instead a dimly lit Machiavellian HUMINT world. However, neither of these two views adequately represents the situation. In the first place, the technophile view which is put forward underestimates the limitations of technical methods, while in the second place the HUMINT view proceeds with a weak understanding of how terrorism, proliferation and serious crime are increasingly using the global infosphere as their operating medium. That intelligence policy-makers must somehow optimize their commitments to human and technical methods may be a truism, but how that balance is to be evaluated is far from a trivial calculation, and what both views of the future underestimate is the interdependency between human and technical methods.

HUMINT Vs TECHINT I: WOOLSEYAN SNAKES ON THE INFORMATION SUPERHIGHWAY

The world of information warfare is, for the most part, a high-bandwidth, high transmission rate universe, and references to espionage and covert action in the literature have, for the most part, reflected that. Martin Libicki of the National Defence University has proposed a sub-class of information warfare that he has termed 'Intelligence-Based Warfare', or IBW. According to Libicki:[IBW] occurs when intelligence is fed directly into operations (notably, targeting and battle damage assessment) rather than used as an input for overall command and control . . .As sensors grow more acute and reliable, as they proliferate in type and number, and as they become capable of feeding fire control systems in real time and near-real-time, the task of developing, maintaining and exploiting systems that sense the battlespace, assess its composition, and send the results to shooters assumes increasing importance for tomorrow's militaries.(Libicki 1995)

However, Libicki's concept is very much confined to military operational intelligence, to the fast-paced high-bandwidth battlefield of the late twentieth and imagined twenty-first centuries, in which the humble footsoldier holds the grandiose title 'warfighter' and Pentagon press releases assure us that high-tech enhances not only lethality and survivability but also 'operations tempo' (Der Derian 1994: 118). Secret intelligence is a very different matter, in which operations are planned and undertaken in terms of months or years rather than minutes and seconds, and the content and sensitivity thereof tend to confine its product to strategic rather than tactical concerns. None the less, a great deal has been made of the potential for cybernetic penetrations to access and disrupt sensitive information systems, with the dangers of 'hacker war' not to be underestimated in an increasingly computer-dependent world. There has also been an accumulating body of evidence that this concern is well founded. Since Clifford Stoll published The Cuckoo's Egg (1990) the world has become well acquainted with the methods employed by Marcus Hess, the 'Hannover Hacker', and his colleagues on behalf of the KGB. However, it might well be argued that the approach of 'hacking into' computer systems is a great deal more limited than might appear in the popular media and some of the more optimistic literature on information warfare. Computer systems can be made secure by physical isolation ('stand alone' systems), or placing them behind hardware defences such as one-way gates and software defences or 'firewalls'. Although Marcus Hess's ring gained access to, according to one estimate, 'fifty military computers at the Pentagon, various defence contractors, the Los Alamos Nuclear Weapons Laboratory, Argonne National Laboratory, the Air Force Space Systems Division and various US military bases around the world (Madsen 1993: 418), perhaps the most distinctive feature of this effort was the very low grade of information acquired. Stoll, the astronomer-turned-sysop who pursued Hess down the telephone lines to Germany, found US Federal Agencies deeply uninterested in pursuing Hess on the grounds that none of the systems attacked held classified or secret information (Stoll 1990: 233 and passim). Indeed, a 1996 report by the US General Accounting Office recently estimated that Pentagon systems had been attacked roughly 250,000 times during 1995, with some 160,000 successful penetrations resulting. The General Accounting Office investigation was brought about by the penetration by a British 16-year-old hacker calling himself the 'Datastream Cowboy', acting under the supervision of a figure using the e-mail handle 'Kuji' whom US authorities suspected of being a foreign intelligence officer. Just as with the case of the Hannover Hacker, however, the Pentagon has taken the reports calmly, noting that no systems handling classified or secret information has been compromised (Walker 1996). In the event, Kuji and the Datastream Cowboy turned out to be far less sinister than US defence officials expected – two British adolescents determined to find evidence of X-files-reminiscent conspiracies and aliens (Campbell 1997). Curious in all of this, however, is the peculiar combination of one group of US defence information security specialists pursuing a perceived threat to the bitter end while another group write such penetrations off as non-threatening because no 'sensitive compartmented information' (SCI) has been compromised.

Unfortunately, it is either naive or disingenuous to suggest that non-classified information has no intelligence value. Much of the information compromised in the US defence computers has been logistical, and this alone can be highly revealing about a nation's defence capabilities and intentions.2 Operation-ally, cybernetic penetrations achieved by those such as Marcus Hess and the Datastream Cowboy occupy a grey zone between open sources and covert collection. Much as train watching may be a legitimate, if esoteric, pastime during peace, it has also traditionally been a major field of HUMINT collection, monitoring the movements of troops and stores in enemy territory, during times of war (see, for example, Landau, 1934 passim or Verrier 1983: 188–9). The acquisition of logistical intelligence becomes the subject of covert collection as soon as it is locked in office cabinets or placed behind user names and passwords.Moreover, in many systems, once one is past the front line defences, the data stored on and electronic correspondence traffic conducted therein is en claire. E-mail, especially, represents a potentially rich vein of raw information, in the same league as telephone and postal interception. If nothing else, such non-classified, so-called 'low-grade' intelligence can provide an analytical background in which context higher-grade intelligence sources may be interpreted and assessed. Regardless of how one might debate the real intelligence payoff of HACKINT to date, the ability to slip past security protocols and snatch passwords used within a system with sniffer programmes has made the development of 'firewalls', overlaid upon a TCP/IP architecture developed originally to promote wider access rather than constrain it, a growth industry.

With all the discussion of the technical collection methods arising out of the new information technology, it comes, therefore, as something of a surprise that in their book War and Antiwar, with its emphasis on high-tech, computer-intensive 'brain force' forms of conflict, Alvin and Heidi Toffler should wheel about and conclude that the prevalent form of intelligence gathering for the 'third wave' would be HUMINT. 'The shift to a third wave intelligence system', they propose, 'paradoxically, means a stronger emphasis on human spies, the only kind available in the 'first wave' world. Only now, first wave spies come armed with sophisticated third wave technologies' (Toffler and Toffler 1995: 186). They do not indicate what third wave technologies will arm these HUMINT sources, one can only infer that they are referring to improved clandestine communications, since they rule out TECHINT methods. Their reasons are that 'the best satellites can't peer into a terrorist's mind', nor into that of Saddam Hussein. In this sense, the Tofflers are repeating an argument for the virtues of HUMINT which has been in play throughout the increasingly TECHINT-intensive Cold War, which is that satellites can show where an adversary's armies are located, but they cannot tell you what he or she intends to do with them. The Tofflers are also falling in line with a view of the changing priorities of HUMINT and TECHINT which has been propagated by intelligence policy pundits since the end of the Cold War. Briefly put, the small number of large-scale threats presented by the Cold War nuclear stand-off have been replaced by a large number of small-scale threats such as terrorism, local wars and serious crime which, while less amenable to imagery and signals intelligence, none the less present a very real need for timely and reliable intelligence. As former DCI James Woolsey has observed, the Soviet dragon may have been slain but the global forest is filled with snakes. A number of authors have therefore suggested that intelligence is headed for a new era dependent upon HUMINT to pursue the snakes of the 1980s (Adams 1994: 311–15; Boren 1994: 55–6).

The shift from TECHINT to HUMINT certainly appears compelling if the proliferation of small-scale, distributed intelligence threats after the Cold War is taken out of the context of the growing global infosphere. In February 1996, during a briefing to British academics under Chatham House rules, a senior UK intelligence officer was asked whether SIGINT had indeed waned in importance as expected. Surprisingly, he responded with a firm denial. 'If anything', he said, 'it is even more important. More terrorists and drug barons are using cellular phones and satellites to talk to each other than ever before.' The lesson here appears to be that the increasing opportunities for technical methods presented by the rapidly expanding, world-wide information infrastructure have more than offset the difficulties of locating and targeting the new threats. Perhaps IMINT may fade in importance, but in its place hacking and communications intelligence are rapidly expanding. During the last decade of the Cold War, the French human source FAREWELL provided documents to the French security service, the Direction Securite Territoriale (DST), indicating that up to 2.4 per cent of the entire Soviet intelligence effort was taken up by cybernetic penetrations alone (Madsen 1993: 419) and this in a society with a far lower per capita availability of the necessary hardware and, skilled operators than the West, and during a period when cyberspace was a far more radically circumscribed place than it is now. Technical methods produce intelligence sui generis; human sources cannot and should not be expected to produce on the same quantitative scale as technical ones. Thus it seems likely that technical methods will remain the large-scale producers, generating the lion's share of the national raw intelligence 'take' for the foreseeable future.

However, information technology may be said to 'giveth with one hand, and taketh away with the other', because the same technologies which promise so much to collection may just as easily deny intelligence collectors that wealth. For there is a sting in the tail of the general availability of inexpensive, personal information technology; and that is the general availability of inexpensive, personal information security.

HUMINT Vs TECHINT II: OF STRONG ENCRYPTION AND THE TWO-LEGGED SPY

It is commonly accepted that HUMINT operations all too often depend upon leads from SIGINT. The wartime Double Cross operation depended enormously on the ISOS breaks against German Abwehr Enigma traffic, while the VENONA decrypts proved valuable tools in detecting the Soviet atom bomb spies during the opening years of the Cold War. However, what is all too often underestimated is the frequent dependency of TECHINT on initial breaks provided by conventional espionage. Perhaps the most famous, and possibly in global historical terms the most consequential SIGINT breaks ever, were the Allied successes against German, Italian and Japanese Enigma machine codes during the Second World War. While most of the decrypts produced in the Allied cryptanalytical efforts may have relied on mechanical innovations such as multiple, parallel Enigma engines, the so-called 'bombes', and the first generation of computers such as Colossus, the ability to develop these techniques initially depended on a human intelligence coup by the French Deuxieme Bureau in the form of the clandestine acquisition of detailed drawings of the German Army Enigma provided by walk-in Hans-Thilo Schmidt (Stengers 1984: 127–8). In 1948 Vienna, it was the reports of a human source in the Austrian postal and telecommunications service that alerted the SIS Head of Station to the fact that telephone lines used by Russian forces in the Soviet Sector passed under the British sector, leading to the tunnelling operations there and in Berlin during the 1950s (Blake 1990: 8–9). Similarly, Soviet COMINT benefited enormously in the 1970s and early 1980s from the KGB's Walker spy ring providing them with current US Navy keys and other cryptomaterials, while a great deal of their most important information about US satellites came from Andrew Daulton Lee and Christopher Boyce (Andrew and Gordievsky 1991: 437–9, 440–2). In the same fashion, it was Soviet rocketry manuals, given to the SIS and CIA by Oleg Penkovsky in Soviet Military Intelligence, that helped photo-interpreters identify Soviet missiles in Cuba as well as in the Asian heartland. Because of this interdependency, Britain's SIS has, since the Second World War, maintained a liaison or 'Requirements' Section representing GCHQ, the UK's SIGINT service, at its headquarters circulating intercepts required for SIS operations, on the one hand, while issuing GCHQ requirements for SIS acquisition of cryptomaterials on the other (Davies 1995). Similarly, the Cold War KGB's foreign operations First Chief Directorate established a separate Sixteenth Department to task rezidentura abroad in support of Sixteenth Directorate SIGINT operations (the Sixteenth Department was responsible for sources like the Walker ring and Britain's Geoffrey Prime) (Andrew and Gordievsky 1991: 440). SIGINT, indeed TECHINT in general, is not a stand-alone collection method, and there are good reasons to believe that HACKINT is likely to be no different from IMINT and SIGINT.

Despite recent successes against DES encryption (Levy 1996: 105–8; Ward 1997b: 19), INFOSEC remains a very real potential limiter in the use of COMINT and HACKINT. The days of negligent sysops leaving factory-issue user names and passwords in place are fast passing, while at the same time firewall development is a fast-growing industry. Libicki has identified the necessarily operational implication of all of this off-handedly, almost by accident. Even though many computer systems run with insufficient regard for network security, computer systems can nevertheless be made secure. They can be (not counting traitors on the inside), in ways that, say, neither a building nor a tank can be. (Libicki 1995)

Although Libicki is undoubtedly overestimating the robustness of INFOSEC techniques, and underestimating the volume of resources which may be directed at a target system by a determined assailant, there can be little doubt that the potential of technical methods against modern information security measures such as 'hacking', intercepting satellite uplink/downlinks and detecting hardware emanations3 have been likewise exaggerated. Apart from the likelihood that it is far easier to design a computationally intractable encryption algorithm than it is to find a way around that algorithm, there are also very good historical reasons to argue that COMSEC and INFOSEC techniques tend to run ahead of TECHINT, and that technical penetrations tend to require a kick start from less sophisticated intelligence methods. Even with the resources of a national SIGINT service, to attack an encrypted system success may prove entirely elusive. Consider the fact that during the Cold War, for ambitious careerist officers within the NSA it was the developing world rather than Soviet Bloc divisions that were considered to be where the real action was. This was because the Soviet machine-based ciphers were almost unassailably strong, and it was more cost-efficient to search the inferior cryptosystems of Soviet Bloc satellites in the Third World for Soviet-originated information echoed in their own traffic (Bamford 1983: xx–xxi; Laqueur 1985: 30–1). Similarly, despite the successes of the various Enigma breaks, certain wartime Axis machine codes were never broken. Thus, although once a system has been penetrated, technical methods may produce vast quantities of raw intelligence sui generis, the initial break is always a delicate matter, all too often resulting from a human operation. What is said about human sources regarding distributed threats is indeed true: human sources very often can go where technical ones may not – but they are most useful when they make it possible for technical methods to follow.

Information technology makes the kind of information security which drives Libicki's doubts about cybernetic penetrations cheap and readily available. A mathematician and programmer acquaintance of the author once developed a relatively strong 8-digit prime public key encryption system for his own uses in the mid-1980s, and promptly made that system available in the public domain via a bulletin board system (BBS). Just as it is joked that there is no such thing as a penniless chemistry graduate in Northern Ireland, there is an enormous potential demand for skilled cryptographers in cyberspace, and not always for legitimate national or corporate purposes. There may be various attempts to control strong encryption it is banned in France, and legally a 'munition' in the United States such constraints are as likely to be effective in cyberspace as similar attempted blocks on the Internet against the Big Book of Mischief or pornography. Indeed, one determined Scandinavian user recently successfully acquired the full paper (rather than the actual software) specifications of the non-export version of the Pretty Good Privacy (PGP) encryption software under the constitutional umbrella of freedom of speech (Ward 1997c). Few things are easier to smuggle than information and where export controls may be vigorously enforced, the development of proprietary strong encryption systems, like the early days of programming in general, has all the possibilities of developing into a fast-growing computer cottage industry in the late 1990s and the next century. What intelligence and law enforcement policy-makers must accept is that the proliferation of information security technology, like that of nuclear, biological and chemical warfare technologies, is a matter of when not if (Scallingi 1995).

The upshot of all of this is that just as information technology may indeed have provided a myriad of new opportunities for SIGINT, as well as the emerging field of HACKINT, it has also provided the stuff of countermeasures against SIGINT and HACKINT in the form of readily available and relatively inexpensive encryption. The most easily accessible back-door to any secure information network is not some quirk in the code or buried programmer's secret entrance, but the people operating it. Human agents can provide their own passwords to their controllers, upload sniffer programmes to provide their controllers with passwords used by others, or simply access systems themselves, download the data and then pass it on to those controllers. Human sysops can be turned, suborned, influenced or bought outright to provide the keys to their forbidden cybernetic city. Coupled with technical operations such as intercepting satellite streams, tapping the landlines and 'hacking in' from outside with agent-acquired legitimate protocols to guarantee entry – HUMINT in fact provides precisely the cutting edge HACKINT and SIGINT are likely to need to continue to be the bulk producer of raw intelligence in a world of 'cheap and cheery' information security. It is precisely a fear of widespread relatively strong cryptosystems for the masses, criminal or otherwise, which has driven the debate over key-escrow encryption, first in the United States over the ill-fated Clipper Chip, and more recently in Europe, with the UK only recently developing its own programme under the auspices of both GCHQ and the Department of Trade and Industry (Communications Electronics Security Group 1997). The debate in America essentially devolved into two schools of thought based on different articles of faith meeting at loggerheads. On the one hand, libertarians feared increasing encroachment into private communications by government, while on the other side, law enforcement agencies were facing the loss of much of the intelligence and prosecutable evidence that communications intercepts have always provided.4 By defeating key-escrow legislation, its opponents have in fact created a new problem, that constitutes a no less pernicious problem of civil liberties: denied access to encrypted communications through a law enforcement access field, or keys held by a trusted third party, law enforcement and intelligence agencies will be forced to recruit more and better placed human sources.

Co-conspirators, accessories, friends, relatives, lovers and spouses. Indeed, as one Canadian national security community official has recently observed, of all the intelligence gathering measures available to domestic intelligence services, the recruitment of human sources is 'oneof, if not the, most intrusive means available' (Whitacker 1996: 284-5). A number of Western democracies have had cautionary lessons that they are not immune to overzealous domestic surveillance. The United States, for example, has legacy of the Hoover years at the FBI and COINTELPRO, while in 1921 the British Secret Service Committee abruptly abolished Basil Thompson's civilian Directorate of Intelligence after only two years of its existence. Their fear was that Thompson's wholesale informant-recruitment in all walks of life was moving towards a 'Continental system of domestic espionage' (Hinsley and Simkins 1991: 6). The recruitment of human sources, moreover, typically does not require the stringent legal controls of judicial or political warrants which are required for communications intercepts or the release of coding keys held in escrow. Thus, it is far from clear whether in defeating Clipper in America civil libertarians have, in the long run, struck a blow for or against the civil liberties they strive
to protect.

HUMINT Vs TECHINT III: SPECIAL OPERATIONS AND (DIS)INFORMATION WARFARE


Special operations, or covert action, cover an extremely wide range of tasks. The major covert action functions include: sabotage, in which facilities are destroyed or incapacitated; disinformation, in which false or misleading information is disseminated with an eye to deceiving a target; influence operations, in which propaganda may be disseminated (deceptiveor not) or clandestine support may be provided to individuals or groups in a political arena with the intention of skewing the results of a decision or an election in a particular direction; disruptive action, in which 'you set people very discretely against one another'; and special political actions on the scale of engineered coups in which combinations of influence,deception and intelligence are used to overthrow one government and replace it with another. All of these tasks rely on good, comprehensive intelligence from all possible sources, overt as well as covert, to be executed effectively.

The potential of information warfare for special operations has probably been the most major single concern in the literature on the subject. Much has been written and said about the threats from logic bombs and computer viruses which disrupt, corrupt or even physically disable information systems (Arquilla 1998; Arquilla and Ronfeld 1996; Devost, Houghton and Pollard 1997; Haeni and Hoffman 1996; Johnson 1997; Libicki 1995; Molander, Riddile and Wilson 1996; Schwartau 1996).7 A great deal has been made of the potential of so-called 'cyberwar', but there is very little evidence that viruses can live up to the hype wrapped around them. To be sure, there have been embarrassing incidents where particularly destructive viruses have been distributed accidentally with brand new 'shrink-wrapped' commercial software and the expensive consequences of the so-called 'internet worm'. In Israel in 1988 a virus was discovered designed to prompt infected programmes to erase all of their files on May 13 of that year, that is, the fortieth anniversary of the 'demise of Palestine' (Kupperman 1991: 92). As pointed out by Kupperman, however, this latter device was little more than a 'weapon of political protest', and the potential impact of individual viruses appears to amount to little more than the cybernetic equivalent of blowing up trains by the Resistance.

The analogy between destructive information technologies and wartime sabotage goes considerably further than just individual actions behind enemy lines. During the Second World War, the sabotage measures undertaken by the European Resistance movements in conjunction with the Allied secret services, in particular the British Special Operations Executive (SOE) and the American Office of Strategic Services (OSS) were closely coordinated to act in support of overt, strategic actions. This was most strikingly so in the case of French groups mobilized in conjunction with the D Day landings in June 1944, attacking German lines of communication and logistical support (see, for example, Stafford 1983: 128).

Likewise, it would seem most effective to combine cyber-sabotage with other political action methods at a theatre or campaign level. There is a general sentiment that this has yet to happen, and much of the literature deals in 'scenarios' and 'simulations'. However, there is good reason to believe that such an information campaign has already occurred in the context of a larger disinformation offensive during the Cold War.

One fact which the literature on the USSR's disinformation and disruptive actions has not taken into account is that by the end of the Cold War the USSR and its Soviet Bloc were the world's largest single source of computer viruses. Of the viruses of known origin (371 types, not including variants; of unknown origin 344), 124 originated from the Soviet Bloc (USSR/Russia and Eastern Europe). That is, of the viruses of known origin in circulation, 33.4 per cent of them originated from the Soviet Union and Eastern Europe. Within the Soviet Bloc, the USSR was the largest single source, followed closely by Bulgaria and then Poland.8 This is despite the fact that the USSR had a considerably lower per capita availability of computer facilities than Bulgaria, let alone the Western states (Bontchev 1995). Although there is no conclusive evidence tying the KGB or GRU to the computer-virus explosion at the end of the Cold War, these figures are none the less highly suggestive. Even if the heavy percentage of Soviet Bloc-originated viruses does indeed represent a coordinated effort rather than being, as Vasselin Bontchev has suggested of the Bulgarian viruses, simply the result of a small number of mischievous or vindictive individual programmers (Bontchev 1995), the fact remains that the presence and wide distribution of viruses throughout the second half of the 1980s amounted to little more than a background noise of nuisance value. The disruptive possibilities of cyber-sabotage may be a genuine cause for some concern, but like the efforts of SOE and the OSS and the various European Resistance movements in 1944, they are unlikely to be an unambiguous war winner, and like the viruses and bacterial munitions of biological warfare there is a very real risk of blow-back. Another potential of modern information technology for deceptive, disruptive and political actions lies in the internet as a means of disseminating information or rather, disinformation. As noted above, much of the data and communications traffic travelling behind the world's firewalls and passwords is enclaire, and politically useful information is not always top secret with a codeword. That background hum of electronic correspondence can all too easily reveal the jealousies and coalitions that drive the organizational politics of a target group or community and can be at least as useful in acting against them as the details of their latest and most secret operational plans. As former SIS officer Baroness Daphne Park observed in 1993 about disruptive action: Once you get really good inside intelligence about any group you are able to learn where the levers of power are, and what one man fears of another . . . you set people discretely against once another . . . They destroy each other, we don't destroy them.(BBC 1993)

This is done, she suggests by way of example, by circulating deceptive information about members or groups within the target group which exploit and exacerbate pre-existing divisions. The same strategy was pursued by the Soviet Union in its 'active measures' disinformation and propaganda campaign during the Cold War by making US and NATO strategy appear threatening to the lives and welfare of Western citizens. This ranged from using various shades of propaganda to suggest that policies such as the Strategic Defence Initiative ('Star Wars') was a direct threat to the survival of Western civilians by increasing the risks of a nuclear conflict (Heather 1987) to fraudulent claims and evidence that US biowarfare research had led to the creation and release of the Human Imunodeficiency Virus (Andrew and Gordievsky 1991: 529–30). The rich stream of e-mail and newsgroup traffic within organizations and communities is an ideal medium from which to glean those divisions and hostilities which disruptive action might exploit. The next problem is to generate plausible deceptive materials and then to disseminate them, and in both cases the 'information superhighway' provides a highly promising medium.

There are two main problems to be overcome in circulating disinformation, either as part of a specific disruptive action or as part of a pervasive 'active measures' style campaign: first, getting the disinformation to the people you want to influence, and second, ensuring that information's credibility will be sufficient to convince them. The internet has vast potential for the first of these problems, but presents real problems in terms of the second. To be sure, it is a communications system with global coverage, and the costs of distributing information over the internet are negligible compared with the breadth of resulting distribution, especially when compared to alternative means. On top of this, digital information production and storage also open up previously unmatched opportunities for the falsification of information. As has already been noted in the RAND report: 'Political action groups and other non-governmental organisations can utilize the internet to galvanize political support . . .Further, the possibility arises that the very “facts” can be manipulated by multimedia techniques and widely disseminated' (Schoben 1995). There have been recent, and telling examples of this kind of action, of which perhaps the most telling were the Malaysian riots that never happened. In this case, a handful of individuals circulated false reports of riots by Indonesian guest workers in the Chow Kit district of Kuala Lumpur by e-mail, reports which circulated faster than they could be countered (New Straits Times 1998), and led to a very real heightening of tensions in a country already shaken by the shooting of illegal Acinese immigrants a few months before.

The very opportunities offered by the internet are, however, its weaknesses. The very ease of dissemination by any group or interest, however ill-informed or extremist, and the very ease of multimedia misrepresentation and digital forgery can potentially make information acquired on the internet a debased currency. Even in the scientific community, the reliability of scientific data available on the Internet has come into doubt, with one survey of internet-using research professionals noting that such information was often flawed because of the following:

Units are frequently omitted;
Transcription errors are often encountered;
This leads to a need to find redundant data;
Very few sources have quality assurance statements;
Few of the Web data sites give the source of the data; and
If they do, data are likely to be copied from outdated sources.
(Wiggins 1996)

Given the limited credibility of materials available on the internet, disinformation distributed by the World Wide Web or by e-mail would have to supported by firm non internet sources to be able to achieve and maintain the kind of credibility that an effective deceptive action requires.Thus the use of cyberspace as a medium of disinformation would be most profitably exploited in combination with other, more conventional human and technical methods.

CONCLUSION: HUMINT, TECHINT AND THE FUTURE OF INTELLIGENCE

It is, therefore, apparent that while technical methods, in particular SIGINT and HACKINT, are likely to be the large-scale providers of national intelligence 'take' for the foreseeable future, this will have to be in the context of a far closer integration with HUMINT. The twin trends of proliferating, small-scale but, as Bruce Hoffman has argued, disproportionately destructive, intelligence targets, e.g. terrorism, proliferation, transnational crime (Hoffman 1996) and a logarithmically expanding global 'infosphere' (information superhighway, cyberspace, call it what you will) mean that no simple, straight-line shift from TECHINT back to HUMINT nor a continuation of TECHINT's Cold War prevalence provides a plausible picture of intelligence gathering in the next century. As argued above, HUMINT's capacity to work around the 'cheap and cheery' INFOSEC systems increasingly available to terrorists, illegal arms dealers and drug barons can provide precisely the window of opportunity for communications intercepts and cybernetic penetrations to continue to act as the main bulk, raw intelligence producers. However, no matter how much one falls back onto HUMINT, analysts and policy-makers have always and will always prefer information from technical sources (Laqueur 1985: 31). Nevertheless, human sources are likely to prove a vital point of entry past the 'cheap and cheery' information security of the 1990s, a HUMINT vanguard finding a path that the more powerful technical sources can follow. What must be kept in mind is that HUMINT operations have very real and often underestimated civil liberties implications in the context of domestic criminal and counter-intelligence operations.No one class of source can be a panacea for intelligence needs in the coming decades. As a result, the issue in allocating intelligence resources in the 1990s and the 2000s is not an either/or of technical versus human methods (indeed, it has never really been so) but how to use them in conjunction with a greater degree of unity, of common targeting and coordination, than has previously been the case.

Philip H.J. Davies
Department of Sociology
University of Reading
Faculty of Letters and Social Services
Whiteknights
Reading
RG6 2AA
UK
lwrdaphi@reading.ac.uk


NOTES

1 As a general rule in British literature, the assorted '-ints' are considered
something of an American affectation, although their convenience (but not, perhaps, their
clarity) has led to a more general usage in the field. The term TECHINT derives from technicalintelligence, and refers to any source of information derived from technical or mechanical means – which virtually amounts to anything not actually constituting a living human source. As noted above, TECHINT can be very loosely be divided into SIGINT, or signals intelligence, and IMINT or imagery intelligence. Within IMINT lie PHOTINT, or
photographic intelligence and 'multi-spectral scanning' (MSS) which employs infrared and
microwave imagery as well as visual wavelengths. SIGINT in turn sub-divides into COMINT,
or communications intelligence (e.g. telecommunications intercepts), and ELINT or electronic intelligence which studies non-communicative emissions such as aircraft radars or missile telemetry. In the following discussion I have, for lack of a better alternative, employed the term HACKINT for intelligence gathered from clandestine, network-based computer access as compared with what is sometimes called 'Network-Based Open-Systems' intelligence or NOSINT, as popularized by figures like Robert Steele of Open Systems Solutions, which relies on overt access to Internet information from open sources (some limitations of which will be addressed below in the discussion of disinformation operations). For good, standard discussions of the various 'int' categories see, for example, Michael Herman (1996), Avram Shulsky (1989), Jeffrey T.Richelson (1989) and Walter Laqueur (1985).

2. A number of commentators have noted this. Clifford Stoll (1990) remarks upon
this, as have journalists such as Walker (1996). Acknowledgements also to James Norminton who made this point regarding an earlier version of this paper.

3 Hardware emanations remain one of the least discussed aspects of information
security, while at the same time representing one of the most substantial single points of
vulnerability in the current information infrastructure. Although the vulnerability of cathode
ray tube (CRT) emissions is reasonably well known (hence the fact that one's password is not
echoed verbatim on a computer screen as is one's user name), the fact is that central
processors generate very considerable electromagnetic field which can be intercepted and interpreted.NATO standards for electromagnetic insulating of information systems are referred to as TEMPESTing, but effective TEMPESTing is expensive and most organizations do not bother with this feature, even if they spend a great deal on encryption and firewall services. A brief discussion of this risk appears in Schwartau (1996: 221–31), and an embarrassing account of ineffective TEMPESTing on CIA-recommended Wang desktop computers recommended to the SIS by the CIA appears in Urban (1996: 256).

4 By comparison, debate in Britain over the TTP Key Escrow proposals issued
jointly by GCHQ's Communications Electronics Security Group and the Department of Trade
and Industry have tended to focus not on civil liberties but on the feasibility and potential
inefficiencies of the plan. See, for example, the CESG homepage (CESG 1997), and the DTI's
'Licensing of Trusted Third Parties for the Provision of Encryption Services' (DTI 1997). A
survey of the issue has also appeared in Ward (1997a.)

5 In Britain's 1985 Interception of Communications Act, the 1989 Security
Service Act and the 1994 Intelligence Services Act, communications intercepts all require
warrants 'signed by the Secretary of State', which in British political parlance means
whichever Cabinet Minister whose jurisdiction covers the location of the intercept, e.g. the
Home Secretary within 'the British Islands' and the Foreign Secretary abroad. British
warranting procedures were subject to some criticism in the 1995 Annual Report of the
Security Commissioner (1995), in which it was noted that while the Intelligence and Security
Services are subject to the legal strictures of the 1985 IOCA, 1989 SSA and the 1994 ISA,
police communication intercepts are only covered by the non-statutory 1984 Guidelines. Reg
Whitacker also notes that Federal Canadian reviews of intelligence have proven consistently
resistant to extending the practice of judicial warrants from communications intercepts to
human source recruitment (1996: 286). Human source selection in Canada is handled by an
interdepartmental Targetting Approval and Review Committee (TARC), while all MI5
operations, human or otherwise, require clearance by the Security Service Legal Adviser, and
SIS operations need clearance from the SIS Foreign Office Adviser.

6 By way of some degree of complication, both sides of the Atlantic tend to use
different jargon, as well as making fine distinctions within the field. Most critically, there is a
clear distinction between what the SIS calls special political actions and the CIA terms covert
action, both of which are strictly political, and special operations (employed by both) to denoteactions which involve paramilitary action. In the CIA, this involved two distinct Divisionswithin the Directorate of Operations, Covert Action and International Activities (formerly Special Operations) (Richelson 1989: 16–17). The equivalent sections of the SIS were defunct by the mid-1970s, with such actions being handled at a geographical rather than central level,and in liaison with either the Foreign and Commonwealth Office or the Ministry of Defence(private information).

7 There is a particular fascination throughout this literature with the issue of
terrorism, or of smaller powers clandestinely disrupting the US military machine and 'critical
national infrastructure'. Arquilla's recent piece in Wired (1998) deals with the danger of what
are known in intelligence circles as 'false-flag' operations by extremist groups, with Devost et
al. (1997) being explicitly and centrally concerned with terrorism. By comparison, Schwartau's
(1996) often alarmist style depends heavily on the idiosyncratic and vaguely melodramatic
notion of the 'information warrior', a non-specific creature composed in various parts of
terrorist, vandal and opportunist. Although Schwartau provides an informative blow-by-blow
description of exactly how information security can be compromised and that often based on
his own practical experience, for the most part much of the literature depends heavily on
disturbing, often worst-case 'scenarios' and simulation exercises of the RAND variety.
L. Scott Johnson's article is of considerable interest since it appeared in the semi-annual,
unclassified version of the CIA's in-house intelligence studies journal Studies in Intelligence.
However, Johnson's 1997 discussion is really yet another threat-assessment of information
warfare, rather than any analysis of the role of information warfare and ICTs in intelligence
policy and infrastructure. Much the same sentiment is developed in DCI John Deutch's
briefing to the US Senate, excerpted in Schwartau (1996: 458–9).

8 This data is drawn from the F-Prot virus descriptions database at
http://www.datafellows.fi/ in Iceland in August 1995. As noted, almost half of the total number of viruses in circulation are of unknown origin. It should also be noted that a small number of the viruses in circulation originated after the Cold War, such as 3APA3A which was found on a university system in Moscow in 1994. It should also be noted that the 1995 version of the F-Prot database distinguished between post-Cold War Federal Russia and the USSR.

9. This is not necessarily a universal risk. China, for example, has concentrated on
developing its information infrastructure on the basis of a highly isolated 'intranet'
programme, rather than the open access adopted by most of the rest of the world, providing
them with a potential national bunker in cyberspace from which to conduct informationwarfare on their neighbours or an inconvenient West with near impunity (Davies 1998).


REFERENCES

Adams, J. (1994) The New Spies: Exploring the Frontiers of Espionage, London: Hutchinson.

Andrew, C. and Gordievsky, O. (1991) KGB: The Inside Story of Its Foreign Operations, London: Hodder & Stoughton.

Arquilla, J. (1998) 'The Great Cyberwar of 2002', Wired 6.02 February.

Arquilla, J. and Ronfeldt, D. (1996) Cyberwar is Coming, available http://stl.nps.navy.mil/~jmorale/cyberwar.html (15 January 1996).

Bamford, J. (1983) The Puzzle Palace, London: Sidgwick & Jackson.
BBC (1993) 'On Her Majesty's Secret Service', Panorama BBC Television, 22 November.

Blake, G. (1990) No Other Choice, London: Jonathan Cape.

Bontchev, V. (1995) The Bulgarian Virus Factories. Available:
http://www.einet.net/galaxy/Engineering-and-Technology/Computer-Technology/Security/david-hull/bulgfact.html(20 December 1995).

Boren, D.L. (1994) 'The Intelligence Community: How Crucial?', Foreign Affairs.

Campbell, D. (1997) 'More Naked Gun than Top Gun' The Guardian,
26 November.

Communications Electronics Security Group, homepage available:
http://www.cesg.gov.uk (5 August 1997).

Davies, P.H.J. (1995) 'Organisational Politics and the Development of Britain's Intelligence Producer/Consumer Interface', Intelligence and National Security 10(4) (October).

Davies, P.H.J. (1998) 'Infowar, Infosec and Asian Security', Asian Defence and Diplomacy 4(9) September.

Department of Trade and Industry (1997) 'Licensing of Trusted Third Parties for theProvision of Encryption Services'. Available http://dtiinfo1.dti.gov.uk/pubs/ (5 August).

Der Derian, J. (1994) 'Cyber-Deterrence', Wired September.

Devost, M., Houghton, B. and Pollard, N. (1997) 'Information Terrorism: Political Violence in an Information Age', Terrorism and Political Violence 9(1) (Spring).

Gibson, W. (1987) Count Zero, New York: Ace.

Haeni, R.E. and Hoffman, L.J. (1996) An Introduction to Information Warfare, available
http://www.seas.gwu.edu/student/reto/infowar/info-war.html (15 January).

Heather, R.W. (1987) SDI and Soviet Active Measures: Mackenzie Paper No 14, Toronto: Mackenzie Institute.

Hinsley F.H. and Simkins C.A.G. (1991) British Intelligence in the Second World War Vol IV: Counter-Espionage and Security, London: HMSO.

Hoffman, B. (1996) 'Intelligence and Terrorism: Emerging Threats and New Security Challenges in the Post Cold-War Era', Intelligence and National Security 11(2) (April).

Johnson, L.S. (1997) 'Toward a Functional Model of Information Warfare', Studies in Intelligence no. 1 1997 (unclassified edition).

Kupperman, R. (1991) 'Emerging Techno-Terrorism', in J. Marks and I. Belaiev (eds) Common Grounds on Terrorism, New York: W.W. Norton.

Landau, H. (1934) All's Fair: The Story of British Secret Service behind German Lines, NewYork: Puttnam.

Laqueur, W. (1985) A World of Secrets: The Uses and Limits of Intelligence, New York: Basic Books.

Levy, S. (1996) 'Wisecrackers', Wired 2.03 March.

Libicki, M. (1995) What Is Information Warfare?, Washington, DC: US Government PrintingOffice, also available http://www.ndu.edu/inss/actpubs/ (19 December).

Madsen, W. (1993) 'Intelligence Threats to Computer Security', International Journal of Intelligence and Counterintelligence 6(4) (Winter).

Molander, R., Riddile, A. and Wilson, P. (1996) Strategic Information Warfare: A New Face of War, Santa Monica: RAND.

New Straits Times (1998) 'Third Suspect Held for Spreading Email Rumours', New Straits Times (14 August).

RAND Research Review Information Warfare and Cyberspace Security (1995), availablehttp://rand.org/RRR/RRR.fall95.cyber (19 December).

Richelson, J.T. (1989) The US Intelligence Community, 2nd edition, New York: Ballinger.

Scallingi, P. (1995) 'Proliferation and Arms Control', Intelligence and National Security 10(4) (October).

Schoben, A. (ed.) (1995) 'Information Warfare: A Two-Edged Sword', Rand ResearchReview: Information Warfare and Cyberspace Security, available:http://rand.org/RRR/RRR.fall95.cyber (19 December).

Schwartau, W. (1996) Information Warfare, 2nd edn, New York: Thunder's Mouth.

Security Service Commissioner (1995) 1995 Annual Report of the Security Commissioner,CMD 2827, London: HMSO.

Shulsky, A. (1991) Silent Warfare: Understanding the World of Intelligence, London:Brasseys.

Stafford, D. (1983) Britain and European Resistance 1940-1945, Toronto: University of Toronto Press.

Stengers, J. (1984) 'The French, the British, the Poles and Enigma', in C. Andrew and D.Dilks (eds) The Missing Dimension: Governments and Intelligence Communities in the Twentieth Century, Chicago: University of Chicago Press.

Stoll, C. (1990) The Cuckoo's Egg, London: Pan.

Toffler, A. and Toffler, H. (1995) War and Antiwar, New York: Warner.

Verrier, A. (1983) Through the Looking Glass, London: Jonathan Cape.

Walker, M. (1996) 'Datastream Cowboy Fixes Pentagon in his Sights', the Guardian, 24 May.

Ward, M. (1997a) 'Coded Message Plan “Too Complex”', New Scientist, 26 April.

Ward, M. (1997b) 'Net Surfers Set Cracking Pace', New Scientist, 28 June.

Ward, M. (1997c) 'The Secret's Out', New Scientist, 6 September.

Whitacker, R. (1996) 'The 'Bristow Affair': A Crisis of Accountability in Canadian Security Intelligence', Intelligence and National Security 11(2) April.

Wiggins, G. (1996) 'Data Needs of Academic Research on the Internet'.available: http://www.indiana.edu/~cheminfo/gw/nist_csanewsl.html (24 January 1997).

 



iCS: visit journals.routledge.com

Routledge

 

    Webcraft by q u a n t u m   d o t   k n o w l e d g e
Copyright ©1997, 1998 quantum dot knowledge