GG only - Open Source (3)

Intelligence for the Twenty-First Century

ALAN DUPONT

Strategic discourse over the past decade has been dominated by a debate over the nature of future warfare and whether or not there is a ‘revolution in military affairs’ (RMA). Supporters contend that developments in military technology, especially precision guidance and high-speed data processing, in conjunction with advances in doctrine and strategy, will fundamentally transform the way in which future wars will be fought and privilege RMA- capable forces in the contest to achieve battlefield dominance.1 Sceptics, on the other hand, regard the RMA as being more evolutionary than revolutionary, and argue that many of the technical advances associated with the RMA do not necessarily presage a paradigm shift in warfare.2

However, all agree that timely, accurate and useable intelligence will be critical to the successful conduct of war in the twenty-first century, perhaps more so than in any previous era.

It is surprising, therefore, how little academic attention has been devoted to the changes that are taking place in the technology, management and integration of the intelligence systems that will underpin any RMA. It is the contention of this article that the transformation of intelligence architectures, particularly in the West, is no less profound than that of the weapons, platforms and warfighting systems they are designed to support and enhance. Moreover, the cumulative weight of the changes in prospect will redefine the way in which intelligence is used and conceived. The old demarcation lines between strategic and operational intelligence and between operations and intelligence, once starkly differentiated, will become more amorphous and blurred. Decision-makers will have better access to intelligence as a result of advances in electronic ‘pull’ technology which have made possible intelligence on demand. Open source intelligence (Osint), while unlikely to supplant traditional intelligence gathering, will enrich and add value to national intelligence databases. There is, however, a downside. Information overload, already a serious problem for intelligence analysts and managers, threatens to diminish the gains from technical improvements in intelligence collection and dissemination. Contemporary manifestations of long-standing policy and doctrinal issues

Intelligence and National Security, Vol.18, No.4 (Winter 2003), pp.15–39 ISSN 0268-4527 print DOI: 10.1080/02684520310001688862 © 2003 Taylor & Francis Ltd.

may further erode the promise of an intelligence-driven ‘knowledge edge’ for those states and military forces able to harness the new intelligence technologies.

In order to articulate and cast light on some of these important trends, this essay focuses on recent and anticipated developments in US intelligence and draws out some of the lessons from the September 11, 2001 terrorist attacks on the World Trade Center and the Pentagon. There are two main reasons for drawing on the US experience and vision.

First, the US is at the cutting edge of modern intelligence technology and management. The emerging national intelligence architecture will be the exemplar for the next generation of intelligence systems operated by the US, as well as its allies and coalition partners.

Second, the ongoing debate in the US over the capabilities, role and management of intelligence has significant ramifications for other nations, both friend and foe alike. Not only is the US the pre-eminent state in the contemporary international system, it also sits at the centre of a web of bilateral and multilateral intelligence relationships that span the globe.

Definitions of intelligence abound, all too often obfuscating rather than clarifying.3 A useful starting point is to define what intelligence is not. It is not merely information or data. Intelligence is information or data which has been processed, evaluated and distilled into a form which fulfils some useful purpose, either to inform policy or, in the case of military conflict, to support operations. For the purposes of this analysis the intelligence process or cycle will be disaggregated into five elements – intelligence collection, assessment, dissemination, use and management.4

COLLECTION – EYES AND EARS IN THE SKY

Intelligence collection is increasingly the function of automated systems mounted on dedicated and multi-purpose platforms which include a variety of ground-based facilities, as well as ships, aircraft and submarines. When US and coalition forces deployed to Afghanistan in search of Osama bin Laden at the end of 2001, they were supported by an impressive array of airborne intelligence collection platforms. They included the E-3 Airborne Warning and Control System (AWACS); upgraded U-2 surveillance aircraft; RC-135 Rivet Joint Sigint aircraft; E-8C Joint-STARS radar surveillance aircraft; Navy EP-3E Aries Sigint aircraft and Navy EA-6B Prowler aircraft. But the US also positioned nearly 50 satellites to support Operation ‘Enduring Freedom’, many of them specifically designed for intelligence gathering, illustrating how critical satellites have become to the collection and dissemination of intelligence.5 Their unique vantage

16 TWENTY-FIRST CENTURY INTELLIGENCE

point in the sky makes satellites particularly well-suited to gathering intelligence in a world ever more reliant on global information networks, smart machines and man-made nuclear, biological and chemical processes all of which display tell-tale signatures that are capable of being recorded or ‘imaged’ from space.

To understand the role that satellites will play in future intelligence collection, it is instructive to look at a representative sample of the satellites currently employed by the US which continues to lead the world in space- based intelligence systems.6 The US deploys four broad categories of intelligence satellites.

First, there are satellites equipped to produce imagery from visible light photographs, radar or reflected infrared emissions. An example of the former is the highly successful KH-11 series of imagery or Imint satellites first launched in 1976 which combine a wide-area, high-resolution photographic capability with real-time transmission of the images produced. The resolution of the cameras in the KH-11 and the follow-on Advanced KH-11/Ikon is so good that intelligence analysts are reputedly able to see, although not read, the licence plates of cars from an altitude of 100 miles.7

Lacrosse satellites use synthetic aperture radars to produce images in all weather conditions.8 Unlike satellites which rely on visible light for their images, radar satellites can see through cloud and at night, but their resolution is generally not as good.

A second constellation of early warning satellites is designed to detect ballistic missile launches. They include the Defense Support Program (DSP) satellites and the replacement Space-Based Infra-red System (SBIRS), which will greatly enhance US ability to detect and intercept ballistic missiles while adding technical intelligence collection and battle space characterisation to the DSP satellite’s early warning function.9 SBIRS High, consisting of four satellites in geosynchronous earth orbit and two sensors in elliptical high Earth orbit is designed to give early warning of ballistic missile launches and track them until the booster rockets burn out. The SBIRS Low system of about 24 satellites in low Earth orbit will then take over tracking the warheads from their point of separation until powerful, X-band ground radars lock on.10

A third category of satellites produces Signals Intelligence (Sigint) and Electronic Intelligence (Elint) by monitoring radio and electronic signals. Throughout the Cold War the US deployed a range of satellites for Sigint and Elint collection which grew steadily in capability, sophistication, weight and cost with each succeeding generation. The most important Sigint satellites over the past three decades have been the Rhyolite/Aquacade, Chalet/Vortex and Magnum/Orion series, stationed in geo-stationary orbit at strategic locations around the globe and capable of monitoring an extensive

17INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

range of signals, from missile telemetry to micro-wave communications. The follow on to the Magnum/Orion, nicknamed Jeroboam by one observer, is over six times heavier than the early Rhyolite/Aquacade satellites and is estimated to cost in the order of US $2bn.11

A fourth class of satellites is equipped with sensors that measure seismic, acoustic, chemical and biological signatures. Known as measurement and signature intelligence, or Masint, these satellites can detect evidence of chemical and biological warfare agents or clandestine nuclear tests. With the proliferation of weapons of mass destruction, and fears that rogue states and terrorists might acquire nuclear, biological and chemical weapons, Masint is almost certain to receive higher priority in resources and funding. The US has launched at least one experimental Masint satellite, code-named Cobra Brass, to measure signatures associated with nuclear proliferation. A lightweight satellite known as Forte, for Fast On-Orbit Recording of Transient Events, has been developed for the same purpose. Forte will be equipped with advanced optical and radio-frequency sensors to monitor the flash and emissions from covert atmospheric tests of the kind thought to have been carried out by Israel and South Africa in 1979.12 Much effort is going into creating an integrated Masint architecture and to improving the timeliness, accuracy and volume of Masint data.13

Prior to the 1991 Gulf War, US Imint satellites were considered to be strategic assets and their product was seldom made available for tactical or theatre-level operations. Exceptions were made, but even then satellite imagery was often days old and highly sanitised.14 Delays in disseminating vital satellite intelligence to coalition commanders during Operation ‘Desert Storm’ and security restrictions on its use fuelled calls for space-based intelligence systems to be reconfigured to better serve the operational and tactical needs of combat commanders and precipitated a major rethink of the structure and priorities of US intelligence.15 Since the Gulf War the US intelligence community has come to accept that intelligence derived from satellites and other national technical means must become more affordable and relevant to the needs of the warfighter and the security challenges of the twenty-first century. This has forced a reassessment of the number and mix of satellites in the US inventory and consideration of alternative collection platforms.

Dedicated intelligence satellites are extremely expensive to build and operate. A KH-11 is estimated to cost around US$800m plus another $300m for the launch vehicle. Even a nation as technologically and financially well-endowed as the US cannot afford to invest in too many billion-dollar intelligence satellites.16 As a consequence, the US is moving towards the deployment of a range of smaller, cheaper and more robust satellites to

18 TWENTY-FIRST CENTURY INTELLIGENCE

augment its highly capable but expensive and less flexible KH-11 and Lacrosse satellites.17

Plans are well advanced to develop a constellation of space-based radar (SBR) satellites that will move the US closer to the holy grail of intelligence – around the clock, all-weather surveillance of the globe. SBR will operate as part of an integrated intelligence ‘system-of-systems’ that will link satellites with manned and unmanned airborne intelligence, surveillance and reconnaissance platforms. Each of the 24 to 48 satellites will cost about $100m. When fully operational, SBR will be able to detect moving and camouflaged ground targets in all weather and terrain and instantaneously transmit the images to friendly forces in the field. Later, improved satellites will be able to identify moving air targets, including possibly stealth aircraft.18 The object is to provide the US and its allies with real-time global surveillance for both strategic and tactical purposes. Importantly, and contrary to past practice, it is envisaged that operational commanders will be able to directly task and in some cases control these space-based intelligence assets.19

Central to the new collection strategy is the deployment of diverse imaging sensors aboard single platforms. In the past secondary Sigint and Elint systems were sometimes carried on board Imint satellites, but in general intelligence satellites had one primary function. Satellites of the future are much more likely to be multi-functional, combining visible-light, electro-optical, hyper-spectral and radar sensors, thereby improving the utility and value of each platform.20 Including radar sensors on an Imint satellite, for example, would benefit an operational commander by ensuring that if the satellite’s pass is obscured by cloud, radar images would still be available. Hyper-spectral sensors break reflected light into different spectral bands which help to penetrate camouflage and to identify the location and composition of objects that would otherwise be hidden from even the most advanced visible light cameras and conventional colour sensors.

Improvements in the collection capabilities of the more expensive satellites through enhanced wide-area imaging and hyper-spectral systems are also being vigorously pursued. The successor to the advanced KH-11 is expected to be able to cover eight times the area of its frame with the same degree of resolution. Current wide-area surveillance by intelligence satellites image only about 500 kilometres either side of the satellite track, which is often insufficient to overlay the full battlefield area.21 Making sense of the Imint collected will eventually be made easier by holograms and lenticular displays that present the data in scaled, three-dimensional images.22 Imint is now expected to do more than count numbers or identify military platforms and the whereabouts of command centres. Contemporary Imint must answer when and how, rather than just how many.

19INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

The intelligence collection capabilities of satellites will be supplemented in the coming decades by a range of highly capable, land-, air- and sea-based systems. In the opening sequence of the film The Empire Strikes Back, the second of the well-known ‘Stars Wars’ trilogy, an imperial intelligence unit dispatches an advanced, all-weather reconnaissance drone to seek out the well-hidden rebel forces on the ice planet of Hoth. Although destroyed by the film’s hero, the drone does its job, alerting Darth Vader’s imperial storm troopers to the rebel’s presence. Such leaps of imagination are no longer just the product of the fertile minds of novelists and screen-writers but are today’s reality. Unmanned aerial vehicles (UAVs), military drones and robots will in future play a major role in providing real-time battlefield intelligence in all terrain and climatic conditions. Equipped with sensors that can see through smoke, cloud and bad weather, they will provide critical intelligence and targeting information, monitor troop and vehicular movements and allow post-strike analyses for a fraction of the cost of satellites and expensive aircraft.

Among the current generation of UAVs in the US inventory, the Predator has already proved extremely valuable as an intelligence collector and hunter-killer in Kosovo and Afghanistan.23 Costing around $10m, the Predator can remain over the target area for around 20 hours, relaying data, Sigint and Imint via satellites directly to the field.24 It has seen extensive service in Afghanistan and was directly responsible for a strike against Al-Qaeda leaders that led to the death of bin Laden’s key lieutenant, Muhammad Atef. Pinpointing a hotel in a remote part of Afghanistan with a Predator’s night vision camera on a moonlight evening in September 2001, US commanders, located in faraway Florida, monitored a meeting of senior Al-Qaeda figures in real time and then used the Predator to direct three incoming F-15 Eagle strike aircraft onto their target. As the F-15 bombs slammed into the hotel, killing Atef, the Predator demonstrated its versatility by destroying several Al-Qaeda vehicles in the hotel car park with its two Hellfire anti-tank missiles.25

Newer generations of UAVs are even more capable than the Predator. The much larger Global Hawk has a 3,000-nautical mile range, can reach altitudes of over 65,000 feet and carry payloads of up to 3,000 lb.26

Comparable in size to the venerable U-2 spy-plane it can be equipped with multiple sensors, including electro-optical, synthetic aperture radar and infrared.27 While Predator searches only for a few signals of interest, Global Hawk is designed to be ‘an electronic vacuum cleaner’, sucking up a variety of signals and emissions.28 The US will probably deploy its first dedicated Sigint Global Hawk in 2004 while some UAVs have already been fitted with Elint suites.29 Global Hawk was rushed into service in Afghanistan and has

20 TWENTY-FIRST CENTURY INTELLIGENCE

played a central role in intelligence collection. If the 1991 Gulf War demonstrated the potential of precision-guided munitions, Operation ‘Enduring Freedom’ in Afghanistan marks the first operational demonstration of the US ability to conduct real time, net-centric warfare featuring UAVs as a core enabler.

HUMAN SPIES

Old-fashioned human spies, producing so called Human Intelligence or Humint, are still crucial to intelligence-gathering notwithstanding the impressive advances in technical collection by unmanned systems. September 11 was a salutary reminder to the US that no amount of technological superiority can compensate for quality Humint. In the aftermath of the successful strikes against the World Trade Center and the Pentagon, the Central Intelligence Agency (CIA) came under sustained criticism for neglecting covert operations and presiding over a decade-long decline in its ability to gather Humint. When the US decided to deploy forces to Afghanistan the CIA had only a single Afghan analyst and a handful of agents fluent in the country’s many dialects. Unsurprisingly, its ability to penetrate Al-Qaeda was virtually non-existent. The CIA has since begun the lengthy task of re-building its clandestine networks, re-learning the value of people and language skills, and recruiting individuals with ‘the savvy to take risks’.30 But these efforts are unlikely to yield results for several more years.

A great deal of the intelligence from agents in the field is derived from, or communicated by, technical systems and devices which can rival satellites and UAVs in their sophistication and stealth. They include state-of-the-art audio and visual eavesdropping equipment, frequency-hopping communications and the use of randomly generated computer encryption. Humint has traditionally been considered a potentially high-value but low- volume contributor to the overall product of Western intelligence communities. However, like other collectors, the tasks and modus operandi of human agents are being transformed by technology and changing priorities.

Clandestine services will increasingly be directed towards support for military operations and the penetration of hard targets, and less for turning out political and economic reports which may add only marginally to information already available from diplomatic and informed public sources. In wars of the future, whether against conventional forces or in low- intensity conflicts, Humint will be vital to the successful conduct of military operations. Even in areas where intelligence-gathering is largely the province of automated collection, humans will play significant roles in deploying and monitoring small intelligence-gathering robots known as

21INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

microbots or insectoids. Insectoids are essentially miniature robots which can look like insects – or anything else for that matter – and can be configured with a variety of sophisticated sensors. Their low signatures and small size making them extremely difficult to find. In a typical deep penetration operation a special force unit might deploy a mix of intelligence insectoids and walk them forward to positions where they could monitor a critical command facility and provide warning of entry and egress by enemy commanders and leaders.31 Alternatively, insectoids or other tiny microbots could be used to determine if an ostensibly commercial building is actually a front for a clandestine chemical or biological weapons programme.32

INTELLIGENCE ASSESSMENT, DISSEMINATION AND

THE POLICY PROCESS

If intelligence collection in the twenty-first century is likely to be dominated by smart machines, intelligence assessments will still reflect the perspicacity of human minds. No amount of raw data can substitute for an insightful human analyst able to discern the critical policy or operational significance of an event, action or trend which may be hidden within a mass of confusing and contradictory information. In attempting to quantify the value of high-quality, finished intelligence former US Secretary of Defense, James Schlesinger, once remarked: ‘when you have good analysis, it’s more valuable than the facts on a ratio of ten to one’.33

While one might quibble with Schlesinger’s arithmetic it is difficult to dispute his basic contention. Major intelligence failures are seldom rooted in lack of information. They are generally failures of analysis and sometimes also of dissemination.34 In the future, they may also result from information overload. The CIA has admitted that the delay in recognising that the Iraqis were storing chemical weapons at Khamisiyah in 1997 was largely due to problems in managing the vast amount of data available to the US intelligence community. And the problem is likely to grow as the volume of raw intelligence multiplies. The US Future Imagery Architecture, for example, is expected to result in a tenfold increase in imagery.35

With the world awash in a sea of information the task of the intelligence analyst is growing daily more difficult and mastery of the art of assessment will be an even more indispensable component of tomorrow’s knowledge edge. To sift through, organise and evaluate the vast amount of classified and unclassified raw information now available will be a major challenge for even the most informed and agile of human minds.36 In the future there will be an even higher premium on analytical skills, particularly at the strategic or national level, where relatively few are sufficiently well-trained or equipped intellectually to deal with higher levels of analysis and the

22 TWENTY-FIRST CENTURY INTELLIGENCE

aggregation of information.37 Analysts who are able to make quick, accurate and informed judgements on fast-moving events and to articulate them in policy relevant assessments will be highly sought after. But the analysts of the future, no matter how gifted or well-trained, are unlikely to realise their full potential unless they can be accommodated in an organisational environment which encourages innovative thinking and allows for genuine debate on issues and problems which may be contentious, ambiguous and sometimes inherently unknowable.

Even here technology is having a significant impact reshaping, in quite fundamental ways, the whole intelligence assessment and dissemination process and redefining the relationship between intelligence producers and consumers. Consider the traditional intelligence cycle which typically begins with a set of broad requirements purportedly representing the collective intelligence needs of customers ranked in priority order. Collectors, both human and technical, work to these requirements, producing raw intelligence reports and data. These are fashioned by analysts into assessments of various kinds and at different levels of classification which are distributed to diverse customers. Economic policy-makers receive forecasts of oil and energy trends, ministers their briefings on the character and peccadilloes of their foreign counterparts and military consumers receive critical data on the capabilities of a potential adversary’s weapons systems or order of battle.

The ‘push’ architecture which supports this edifice requires the analyst, not the user, to select from the available information what he or she believes the consumer wants to know. Further up the line intelligence managers and coordinators determine which assessments will be produced and when. Frequently, however, the product does not meet the consumer’s needs, either because users are ignorant of what is available or the producer is unable to determine precisely what it is that the user really wants. The value of the finished product can be further diminished by delays in dissemination, especially of national estimates and assessments that may take weeks, and sometimes months, to prepare and endorse.

Since the late 1980s these deficiencies have been magnified by revolutionary changes in information and communication technology, especially the advent of commercially available, real-time information services on television, cable news networks and the Internet. When the Gulf War exploded in 1991, military and civilian officials alike were compelled to watch television images of terrain-hugging cruise missiles meandering down Baghdad streets while the equivalent satellite imagery was not available to decision-makers for hours and frequently days. This same revolution, however, now enables consumers to access the intelligence they want on demand, through what has been dubbed ‘pull’ architecture, so

23INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

called because users electronically pull down or download intelligence from a networked database through a dedicated terminal.38

Furthermore, it is now possible to access ‘all-source intelligence’ from a computer terminal that integrates classified and unclassified information. Thus policy-makers with the appropriate security clearances have the considerable advantage of being able to select and download material from the same raw and finished intelligence product that is available to the professional intelligence analyst. And they can do so at a time of their own choosing. This clearly has major implications for the producer–consumer relationship.

On the positive side, the value and relevance of intelligence to policy- makers is likely to be significantly enhanced by the new pull architecture which, in theory at least, should stimulate more informed discussion of issues and allow consumers to quickly differentiate useful, high-quality analysis from pedestrian, low-value-added assessments or largely descriptive summaries of raw material. In a process known as ‘disintermediation’, top-level decision-makers will tend to bypass middle- level managers and coordinators in favour of downloading their own data or speaking directly to the geographical or functional specialists.39

Disintermediation will reinforce the trend towards flatter management structures. In the longer term a more discriminating and informed customer community will demand higher product standards forcing producers of intelligence to become more responsive and to subject their assessments to the marketplace of competitive analysis.

If national intelligence communities fail to meet these higher standards they risk marginalisation. One major risk of the new pull architecture is that policy-makers with direct access to a comprehensive intelligence database may take it upon themselves to act as their own intelligence analysts, either through hubris, dissatisfaction with the existing service or because of time constraints. The temptation to do so will be even more acute if the professionals are demonstrably unable or unwilling to deliver a satisfactory product. Over 25 years ago, the Church Report of the US Senate highlighted the danger of do-it-yourself intelligence analysis. The report warned that not only may consumers of intelligence ‘be depriving themselves of the skills of intelligence professionals; they may also be sacrificing necessary time and useful objectivity’.40

Technological change has sharpened another long-standing and related debate over the desirable degree of separation between the policy and intelligence communities. Managers of intelligence tend to argue that the two processes ought to be separate and even compartmentalised. The reason? ‘Analytic intelligence provides a reality check on policy-driven hopes and aspirations. It fosters objectivity… Policy-makers, optimists by

24 TWENTY-FIRST CENTURY INTELLIGENCE

nature, are apt to clutch at straws and tiny crevices as they scale impossible heights to reach desired goals.’41 Policy-makers respond that such distinctions are over-drawn. While they agree that intelligence officials should refrain from offering policy advice, they dispute the claims of intelligence officials to a monopoly on objectivity and impartiality and their attribution of automatic policy bias to the policy-maker. US Deputy Secretary of Defense Paul Wolfowitz is one who believes that neither side can completely avoid policy bias when it comes to dealing with uncertainty.42

Both views clearly have merit. Those intelligence officials who cross the policy line do so at their own peril, risking their objectivity and credibility. At the same time intelligence analysts must accept that their assessments and judgements are not completely free from intellectual and policy bias. But this dichotomy obscures the more important question of how producers and consumers can better understand each other and work more closely together in the interests of constructing an effective system of intelligence support for policy and operations.43

Decision-makers ought to recognise that a competent intelligence analyst is uniquely equipped to provide meaning, clarity and order to the rising tide of information which confronts and informs them. The intelligence community, on the other hand, must customise and render more user-friendly the material and support which it provides, particularly to politicians and senior officials whose culture is primarily an oral not a written one.44 Intelligence analysts must also be more accessible to decision- makers and prepared to dialogue regularly and freely with experts and thinkers outside their own somewhat cloistered environment. Excessive secrecy and insularity, as Gregory Treverton observes, is the enemy of timely and insightful intelligence.45

NEW INTELLIGENCE TARGETS AND PRIORITIES

Intelligence planners will be compelled to broaden and diversify the scope of their collection and assessment activities in response to a range of new security threats such as terrorism, transnational organised crime, illegal migration and environmental degradation.46 These non-traditional security threats have intruded onto the international security agenda with growing frequency over the past decade and they are beginning to be reflected in national intelligence priorities.47 In 1995, the then US Vice President Al Gore instructed the CIA to evaluate for the first time the degree to which environmental factors influence global and national security.48 Counter- terrorism, long a priority for police forces and domestic security agencies is a task which is increasingly likely to fall within the purview of the military

25INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

and to warrant the allocation of strategic and operational intelligence assets. Much of the existing apparatus of intelligence can be readily adapted to

new targets and purposes while intelligence analysis and collation is one of the most immediately transferable of skills. There is also political advantage in making intelligence more available and relevant to the community at large. Domestic political considerations were important in the decision by the Clinton administration to de-classify and release some 800,000 satellite images in response to requests from the scientific community for help in monitoring the environment.49 In the longer term, however, the allocation of scarce intelligence resources to such tasks will have to be carefully weighed against the needs of other users and be consonant with the future role and objectives of national intelligence communities.50 For defence forces judgements will have to be made about the priority to be given to Operations Other Than War (OOTW) and the extent to which new capabilities are needed to counter transnational threats such as drug trafficking, terrorism and people smuggling.51 In an era of holistic intelligence these issues need to be carefully considered when developing national intelligence collection and assessment priorities.

OPEN SOURCE INTELLIGENCE (OSINT)

Another important trend with potentially far-reaching implications for intelligence professionals and their diverse constituencies are the qualitative improvements and volume growth in Osint. Estimates of the proportion of US intelligence derived from non-classified, publicly available sources ranges from 40 to 95 per cent, although a commonly accepted figure is 80 per cent.52 While the balance between classified and non-classified material is difficult to quantify precisely, for many subjects there is little doubt that Osint is at least as important as classified intelligence. Most intelligence professionals accept that Osint should be better integrated into the national intelligence process but some want to go much further. One well-known advocate, Robert Steele, believes that Osint ‘can serve as a foundation for reinventing and reorienting the clandestine and technical disciplines’.53

While such claims may overstate the role that Osint can play in the intelligence architecture of the twenty-first century Osint clearly has much to offer. It is a relatively cost-effective way of tapping into the expertise that exists in the community at large thereby multiplying the pool of available intelligence and freeing up scarce resources for tasks where Osint is less able to contribute. Osint is often useful to policy-makers precisely because it is unencumbered by security caveats and restrictions on dissemination. Osint can also be tailored to the tastes and interests of individual consumers at minimal cost which intelligence agencies find difficult to emulate. There

26 TWENTY-FIRST CENTURY INTELLIGENCE

are already in existence a significant number of reputable private organisations capable of providing timely, high-quality political, economic, strategic and technical analyses on a subscription basis. Some well-known examples in the strategic domain are the various reports and assessments produced by the UK-based Oxford Analytica, Jane’s Information Group, the International Institute for Strategic Studies and Stratfor.

Osint is also set to challenge the major powers’ dominance of space- based intelligence systems. By 2015, most small and medium-sized states will be able to acquire intelligence from a diverse range of commercial satellites that are steadily closing the once considerable technological gap with dedicated intelligence satellites.54 The proliferation of commercial satellite technology, especially the order of magnitude improvements in high-speed data transmission and imagery resolution, are making it feasible for virtually all states to integrate commercial systems into their national intelligence network. Although currently lacking the resolution for precise targeting information, off-the-shelf images from Landsat and SPOT satellites have provided useful area coverage and hyper-spectral imagery for over a decade freeing up military satellites for high value tasks. SPOT images were used by coalition commanders for planning air raids on Baghdad during Operation ‘Desert Storm’.

The next generation of commercial satellites will be far more useful for targeting and intelligence purposes because they are capable of producing relatively high-quality images of less than one metre, width panchromatic (black and white) resolution. A resolution of one metre is sufficient to identify aircraft and ships, and to classify vehicles, making it more difficult to conceal military capabilities and preparations for war. In October 2001, the QuickBird observation satellite operated by DigitalGlobe was launched, further narrowing the gap with US spy satellites. QuickBird is capable of panchromatic resolutions of 61 centimetres and 2.5-metre multi-spectral images.55 Data collected by scientific spacecraft and satellites can also yield valuable intelligence. Radiometer data from NASA’s Modis and Sea WiFS spacecraft have helped coalition special forces operating in Afghanistan to determine the specific characteristics of snow, sand and dust thereby aiding them with tactics and targeting.56

However, Osint has its dangers and limitations. By definition commercial imagery is available to anyone who is prepared to pay, including rogue states and terrorist groups. That is why the US has developed contingency plans to deny potential enemies access to high- resolution commercial imagery. Immediately after the September 11 terrorist attacks, for example, the US National Imagery Mapping Agency (NIMA) banned the release of 3D terrain elevation measurements made over the US by the NASA/NIMA shuttle radar topography mission.57 More

27INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

generally, Osint varies tremendously in quality and reliability. Because it is driven primarily by commercial considerations Osint tends to focus on topical issues which may be ephemeral rather than fundamental. Underlying trends or subjects that are considered to be too technical or arcane are less likely to merit commercial attention and investment when in fact they may be vital to national security planning.58

Ironically, the phenomenal growth in Osint over the past decade is a major reason for the information overload which intelligence assessment agencies are struggling to overcome. The CIA estimates that the amount of open source information grew by a factor of ten in the five years between 1992 and 1996, a rate that shows no signs of slowing.59 The key question for intelligence communities is not whether Osint has an important role to play but how can it best be integrated into national intelligence systems.

HOLISTIC INTELLIGENCE

Fusing and integrating previously unlinked platforms, technologies and resources into holistic intelligence systems will be a defining trend of the next decade, and effective management of holistic intelligence may be the key to unlocking the promise of the still nascent RMA. Managing the vast amount of information gathered, analysed and disseminated by the military will be the critical force multiplier of the future, equivalent in its impact to stealth technology. The shape of tomorrow’s intelligence architecture is already discernible. Secure intranets, adapted to commercially available software and Web-systems, which find, organise, filter and analyse information are now in common use. Classified and unclassified material, including imagery, Sigint, Elint and Masint, are available literally at the touch of a key. Small, hand-held personal digital assistants that give soldiers in the field access to huge databases using web-enabled, wireless communications, are transforming the use of intelligence at the operational level of war. Advanced search engines and text analysis tools like Pathfinder are having a similar effect in the strategic domain, allowing analysts to swiftly extract useable intelligence from large amounts of data.60

The US Defense Intelligence Agency plans to operate a virtual intelligence network, known as the Joint Intelligence Virtual Architecture (JIVA) which is intended to establish a single ‘knowledge’ database for commanders, policy-makers and those responsible for acquisition. JIVA will link geographically separated users and producers of intelligence in a global electronic network featuring full motion video and graphics displays, interactive data manipulation and search and retrieval from mass storage.61

The US military has made steady progress towards its aim of integrating the capabilities of multiple intelligence collection platforms to produce a

28 TWENTY-FIRST CENTURY INTELLIGENCE

comprehensive picture of the battlefield in all its complexity regardless of weather, terrain or time. For example, the Joint Forces Air Component Commander Situational Awareness System (JSAS) feeds intelligence from satellites, aircraft and UAVs onto a single screen showing friendly and enemy forces in unprecedented detail and clarity.62

Established on 1 October 1996, the National Imagery and Mapping Agency (NIMA) provides a ‘one-stop shop’ for imagery, mapping and dissemination services for commanders and decision-makers, subsuming the functions previously carried out by four separate agencies.63 Senior Air Force commanders now talk about ‘predictive battlefield awareness’ where superior intelligence and profiling will provide the capability to actually anticipate an enemy’s actions and respond accordingly. In a typical pre- planned attack based on predictive intelligence, a loitering stealth fighter armed with a new generation of small, precise stand-off weapons waits for the tell-tale glint of metal under camouflage that tells the pilot his quarry is moving out of an underground bunker. As anticipated, the target emerges only to be destroyed within seconds by a missile launched from the waiting fighter.64

What is often not appreciated is the massive amount of data modern weapons need for maximum performance. Smart weapons such as cruise missiles and precision-guided munitions require high-resolution imagery and accurate positional data to be effective.65 Some Cold War strategic intelligence systems have been adapted to meet these needs but a new generation of purpose-built intelligence and targeting systems designed to support operational commanders is beginning to emerge. In early 1998, the US Air Force deployed an initial version of the Rapid Targeting System (RTS) to Kuwait. Its ground-based computer system has been designed to provide real time imagery from a range of photographic and radar satellites for strike aircraft. These images are correlated with archived intelligence obtained from airborne reconnaissance platforms and matched against a map to give the highly precise GPS coordinates essential for accurate targeting.66 A complementary system is the Moving Target Indicator (MTI) used by Joint Surveillance Target Attack Radar System (JSTARS) aircraft. MTI has a unique capacity to track and target moving vehicles through its synthetic aperture radar which is able to see in all weather conditions including through obscurants like smoke.67

The fusion of intelligence with command, control, communications, computers, surveillance and reconnaissance (C4ISR) will blur the distinction between each of these functions and compel commanders and planners to treat them as an integrated whole with the end goal being the creation of ‘seamlessness’ across all elements of a fighting force.68 The Command and Control (C2) Augmentation System that was deployed to

29INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

Bosnia indicates the potential of the integrated C4ISR architecture of the future. The C2 Augmentation System handled a thousand times more communication traffic than was possible during the 1991 Gulf War and merged data from many different sources, routing the information electronically to wherever it was needed. It also allowed other nations to share the same intelligence as US forces, or a sanitised version, depending on their status within the hierarchy of US allies and coalition partners.69 By 2015, the new C4ISR architecture will give the US and its allies an unmatched ability to identify, target and destroy hostile forces and war- fighting infrastructure.

The wider dissemination of fused intelligence will not only force intelligence communities to become more user-friendly and less production-oriented, it will also lead to new divisions of labour and the decentralisation of decision-making, particularly in the military realm. Traditionally, sensors and shooters have belonged to the one organisation. Ground attack aircraft were directed to their targets by military forward air controllers or by self-contained, highly specific sensors attached to the aircraft or ordnance being delivered. However, today’s generic intelligence platforms and sensors feed their data into a wide variety of weapons systems and the controllers of the sensors are often from different bureaucracies to the shooter’s. Holistic intelligence will empower users at all levels, further blurring the boundaries between tactical and strategic level intelligence and creating more decentralised intelligence systems that are agile, network- based and highly responsive to the needs of its users.70

WILL THE ‘FOG’ OF WAR BE LIFTED?

However, for all the undoubted improvements in the collection, integration and management of intelligence, either extant or in prospect, there are serious reasons to doubt whether the fog of war can be sufficiently penetrated or lifted to permit the ‘dominant battlespace knowledge’ which RMA enthusiasts proclaim.71 Intelligence, especially in war, will remain a less than perfect approximation of an adversary’s intentions and capabilities. Good intelligence can reduce uncertainty and inform decision- making. It can help distinguish vital signals from background noise, but it is subject to the same foibles and fallibilities as its human producers and users. More information does not equate to better intelligence. It may, in fact, diminish the capacity of users to perceive, know and predict. The Jeremiah Report into the failure of the US intelligence community to provide warning of the Indian nuclear test on 11 May 1998 found that US intelligence satellites produced far too much information for overworked and under-trained analysts to handle.72

30 TWENTY-FIRST CENTURY INTELLIGENCE

There is also the problem of what John Ferris and Michael Handel call the ‘Schwarzkopf syndrome’ – ‘the desire to wait just one more moment in order to read just one more report, the reluctance to act on imperfect knowledge because it is known to be imperfect and that at any point another report might well produce perfection’.73 While this is not just a contemporary failing, the centrality of intelligence to decision-making at both the strategic and tactical level and the sheer volume of information now available can induce a form of intelligence dependence which hinders decisive action. It is not only the volume and detail of information which threatens to overwhelm analytical capability. The rapidity with which information can be collected and disseminated to decision-makers in the modern era poses problems of its own. There is simply less time for reflection and considered judgements.

In future crisis situations strategic analysts will not only have to compete with the near real time images of CNN and other global news services. Their monopoly over the flow of assessed and raw intelligence is also under internal challenge – by collection agencies which make available to key policy-makers raw material which is judged to be of particular importance or time sensitivity; and by individual collectors in the field who use email, the Internet and cellular telephones to disseminate information on fast moving events directly to users. By their very nature such informal intelligence reports circumvent the established checks and balances within the intelligence system and raise issues of control, accountability and authenticity. Thus, qualitative technical improvements in collection and dissemination can actually undermine the efficacy of the overall system.

On the other hand, the development of ad hoc and informal intelligence networks is not intrinsically undesirable or necessarily a bad thing. In some circumstances the fluidity and flexibility of such networks may be preferable to the relatively unresponsive and de-personalised ‘coordinated’ intelligence that is still the stock trade of assessment agencies.74 Much depends on the balance struck between the two. Clearly, users will continue to demand more timely and tailored intelligence. But without a mechanism to ensure quality control, informal reporting from the field or unevaluated raw intelligence may prove to be more risky than beneficial.

Militarily weak states and non-state actors such as terrorists and transnational criminal organisations will almost certainly make use of asymmetric strategies designed to combat, neutralise and disrupt the intelligence systems upon which the US and its allies depend for their knowledge edge and combat superiority. Osama bin Laden knew that his satellite phone could easily be monitored and limited its use to non-sensitive matters. Important operational directives were usually sent by courier or trusted personal messengers.75 Asian terrorists and criminals make extensive

31INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

use of the traditional underground banking system (UBS) or ‘hawalla’, as it is known in South Asia. The UBS is a highly personalised, family-based network of gold shops, trading companies and money exchanges which maintains minimal records and relies on mutual trust between brokers for its efficiency and internal discipline. Financial exchanges typically rely on coded messages, telephone calls, handwritten chits or symbols which guarantee a high degree of personal security and virtual anonymity.76 Thus, large sums of money can be moved expeditiously all over the world with very little chance of detection by the authorities because they are outside the regulated banking system.

At the technical level there remain many problems to overcome. The exponential increase in high-speed communications and data transfer, particularly of images, has raised concerns about existing bandwidth capacity, the power demands of sustaining the supporting architecture for modern C4ISR systems and the vulnerability of these systems to power failures and hostile action.77 While communications advances such as cell phones and email have opened up new avenues for exploitation, other developments have had the opposite effect. Fibre optic cables do not give off electronic emanations and commercially available high-grade encryption technology and radio frequency hopping has impeded intelligence gathering by severely reducing or complicating technical collection to the point where some potential adversaries may be on the verge of denying access.78

The more reliant military forces become on satellites and electronic data for intelligence and targeting, the more likely it is that potential adversaries will seek to exploit their vulnerabilities. Concealing high value facilities from the prying electronic eyes of airborne and space-based intelligence collection platforms, suppressing electronic emissions or moving potential targets around prior to an anticipated strike are some obvious defensive counter-measures. Offensive measures might include physically attacking critical intelligence nodes, or else rendering them dysfunctional through computer viruses and other forms of information warfare (IW).79

INTELLIGENCE AND INFORMATION WARFARE (IW)

As a result, national intelligence communities will inevitably become more directly involved in planning and executing IW strategies. Intelligence collection agencies will work closely with operational planners to develop and implement IW doctrine and information operations (IO) while assessment agencies will be more actively tasked to report on the intelligence vulnerabilities of potential adversaries and to identify IW targets. Intelligence assets and platforms which are currently used for

32 TWENTY-FIRST CENTURY INTELLIGENCE

monitoring the C4ISR systems of hostile states can also be used more aggressively to help disrupt and corrupt these systems or to assess the results of an IW strike. Rather than shooting down a commercial satellite, which is being used by an adversary for reconnaissance and intelligence purposes, it might be considered preferable to commandeer the satellite by sending a strong over-ride signal ordering the satellite to shut down or change its orbit.80 In IW, operations and intelligence may become so interconnected and entwined that the distinction between the two becomes largely academic.81

IW poses important policy and doctrinal issues for intelligence communities. For example, during a conflict is it better to monitor or destroy an enemy’s C4ISR systems? If the latter, should the task be accomplished with iron bombs or electronic bullets? Of course, this is not an entirely new conundrum. On one side of this long-standing debate are intelligence officials who for bureaucratic, security and cultural reasons are disinclined to actively interfere with systems that have been penetrated and are providing rich seams of intelligence. Military commanders, on the other hand, are more disposed to disrupt or destroy intelligence systems especially if they are causing significant casualties and problems for friendly forces. This dilemma has become progressively more acute as the salience of computers and information systems to modern warfighting has grown, forcing military planners to contemplate a range of aggressive IW options for corrupting, manipulating, deceiving and destroying an adversary’s intelligence systems.82

FUTURE TRENDS

In the decades ahead, the forces of change will continue to play out their dialectic, reshaping and transforming the art of intelligence in a myriad of ways. While information will become more plentiful and less of a privileged source in the ‘global information environment’ of the twenty-first century, paradoxically the demand for timely, high quality strategic and operational intelligence will intensify rather than diminish.83 The ever-increasing volume of information will be a test of a different order for the analysts, warfighters and policy-makers of the future. Yet the core of the intelligence dilemma will remain – how best to reduce the element of risk for decision- makers and cast light on what might otherwise be unknown?

Dedicated intelligence collection platforms will of necessity become multi-functional and the chariots of war will see, hear and sense as well as kill. In the drive to reduce costs, and to extract more return from expensive intelligence assets, commercial off-the-shelf technology and Osint will supplement, strengthen and sometimes replace the classified systems that

33INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

dominated intelligence collection during the Cold War. However, although classified material will make up a smaller proportion of the overall intelligence available to users it will still provide critical pieces of the intelligence puzzle. Just how critical will depend in large measure on the particular puzzle and the urgency of the required response. Classified collection methods are generally more useful for current assessments and for sensitive and difficult to obtain military or technical intelligence rather than long-term forecasts and economic analysis.

Non-state actors will pose challenges of a different order. Operating in the shadows, and steeped in the arts of deception and disguise, terrorists and criminals may be much more difficult to identify, locate and eliminate than the political leadership or military forces of hostile states. Costly, high-tech intelligence systems designed for conventional warfare or monitoring the electronic environment may be ineffective against organisations employing simpler methods of clandestine communications as Osama bin Laden demonstrated in his well-coordinated, devastatingly effective attack against the US that some have labelled this century’s ‘Pearl Harbor’.84 Despite impressive advances in sensors, automation, technical collection and code- breaking, it remains the case that humans provide the critical intelligence edge. There is simply no substitute for effective managers, prescient analysts, gifted linguists and dedicated spies.

The traditional intelligence cycle clearly has less explanatory and organisational utility in the post-Cold War world. The discrete functionality implied in the separation of the intelligence process into collection, collation, analysis and dissemination reflects the concepts, practice and organisational dynamics of an earlier era. What will distinguish the successful practitioners of twenty-first-century intelligence is the ability to fuse and integrate all elements of the process to provide seamless support for policy-makers and operational commanders. As has already been argued, the bureaucratic walls which once separated the providers and users of intelligence are coming down at a rapid rate, removing the artificial and dysfunctional distinctions between strategic, operational and tactical intelligence assets. The imperatives of modern warfare, particularly IW, will reinforce the need for holistic intelligence and erode the distinction between operations and intelligence.

On balance, these developments are likely to enhance the contribution which intelligence makes to national security although they are not without their risks and liabilities. Prudence and historical experience suggest that wise leaders and commanders will retain a healthy degree of Clausewitzean scepticism about the capacity of intelligence to completely eliminate the difficulties and uncertainties involved in attempting to divine the intentions and capabilities of others, especially in the heat of battle. Asymmetric

34 TWENTY-FIRST CENTURY INTELLIGENCE

warfare, human fallibility, technological failure and the development of effective countermeasures will ensure that no intelligence system, no matter how efficacious, will ever be able to completely dispel the fog of war.

NOTES

1. Typical of the ‘true believers’ in the revolutionary nature of the latest generation of military technology is the former vice chairman of the US Joint Chiefs of Staff, Admiral Bill Owens. See, e.g., Admiral William Owens, ‘The Emerging System of Systems’, US Naval Institute Proceedings 121/5 (May 1995).

2. Lawrence Freedman, The Revolution in Strategic Affairs, Adelphi Paper 318 (London: OUP for IISS April 1998) p.8; and Colin Gray, The American Revolution in Military Affairs: An Interim Assessment, Occasional Paper No. 28 (Hull: Strategic and Combat Studies Inst., Univ. of Hull 1997).

3. For a comprehensive discussion of definitional questions and issues, see Angelo Codevilla, Informing Statecraft: Intelligence For A New Century (NY: The Free Press 1992) pp.3–47.

4. There is much debate in the literature about the relevance and adequacy of the traditional intelligence cycle. Among the more insightful works dealing with this issue are Bruce Berkowitz, ‘Information Technology and Intelligence Reform’, Orbis 41/1 (Winter 1997) pp.109–11; and Amos Kovacs, ‘The Uses and Nonuses of Intelligence’, Center for International Security and Arms Control Report (Inst. for Int. Studies, Stanford Univ. Oct. 1996), p.1.

5. Just how important is detailed in Jeffrey Richelson’s pathbreaking book, America’s Secret Eyes in Space: The U.S. Keyhole Spy Satellite Program (NY: Harper & Row 1990). See also Desmond Ball, ‘Desperately Seeking bin Laden: The Intelligence Dimension of the War Against Terrorism’, in Ken Booth and Tim Dunne (eds.), Worlds in Collision: Terror and the Future of Global Order (London: Palgrave Macmillan 2002) pp.66–7, for an analysis of the airborne and satellite intelligence collection systems deployed to Afghanistan in Sept. and Oct. 2001.

6. Indeed, in many areas, the US has extended its lead in space-based systems. Barbara Starr, ‘USA Sees Vital Role for Commercial Satellites’, Jane’s Defence Weekly, 30 July 1997, p.21.

7. Jeffrey T. Richelson, ‘High Flying Spies’, The Bulletin of the Atomic Scientists, Sept./Oct. 1996, p.53; and Alan Dupont, ‘The US and Verification of SALT II’, Pacific Defence Reporter, Oct. 1979, p.26.

8. Desmond Ball, The Intelligence War in the Gulf, Canberra Papers on Strategy and Defence 78 (Canberra: Strategic and Defence Studies Centre, Australian National Univ. 1991) p.15.

9. Statement by Paul G. Kaminsky, Undersecretary of Defense for Aquisition and Technology, to the National Security Subcommittee, House Appropriations Committee, 23 March 1995, <www.defenselink.mil/speeches/index.html>.

10. SBIRS and its functions are described in Richard J. Newman, ‘Space Watch, High and Low’, Air Force Magazine, July 2001, pp.35–8. Satellites launched into geosynchronous orbit 35,000 kms above the equator are able to hover over a single spot because their orbit speed matches that of the earth’s rotation.

11. Desmond Ball, Signals Intelligence in the Post-Cold War Era: Developments in the Asia- Pacific Region (Singapore: Regional Strategic Studies Programme, Inst. of Southeast Asian Studies 1993) pp.15–19.

12. Richelson, ‘High Flying Spies’ (note 7) p.53. 13. ‘Streamlining MASINT’, Aviation Week & Space Technology, 7 Sept. 1998, p.55. 14. During Operation ‘Desert Storm’, frustrated by his inability to locally access vital imagery

which he needed for operational planning, one US Marine Corps General in the Persian Gulf dispatched a personal aide to Washington to pick up photographic intelligence and bring it back by safe-hand.

35INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

15. Photographic imagery took up to four days to reach operational commanders during ‘Desert Storm’ according to the deputy director of the Defense Intelligence Agency, Admiral Ted Shaefer. Shaun Gregory, Command, Control, Communications and Intelligence in the Gulf War, Working Paper No. 238 (Canberra: Strategic and Defence Studies Centre, Australian National Univ. 1991) p.8. Various programmes were already underway in the US well before the Gulf War to make intelligence derived from satellites and other ‘national technical means’ available to operational commanders. There is little doubt, however, that the Gulf War experience accelerated the trend.

16. Bill Sweetman, ‘Spy Satellites: the Next Leap Forward’, Jane’s International Defence Review, Jan. 1997, p.27.

17. Each kilogram of weight adds about US$5,000 to the cost of a commercial launch. Initial enthusiasm for small satellites in the commercial sector was dampened in the 1990s because of a series of high profile failures but they are back on the agenda. Small satellites generally weigh less than 100 kilograms and may in future be launched for as little as $1m. ‘Incredible Shrinking Satellites’, The Australian (Cutting Edge), 16 Oct. 2001, pp.1, 4.

18. Robert Wall, ‘New Space-Based Radar Shaped by SBIRS Snags’, Aviation Week & Space Technology, 18 Feb. 2002, p.30; and Michael Sirak, Jane’s Defence Weekly, 22 Aug. 2001, p.8.

19. Warren Ferster, ‘U.S. Wants Radar Satellite’, Defense News, 16–22 March 1998, p.20. On small satellites, see Pat Cooper, ‘U.S. Intelligence Reshapes Techniques, Goals’, Defense News, 12–18 Aug. 1996, pp.8, 18.

20. For example, an advanced KH-11 satellite launched on 28 Feb. 1990 also carried a Sigint package for intercepting video transmissions and monitoring radio and microwave transmissions. Ball, Signals Intelligence In the Post-Cold War Era (note 11) p.15. In general, however, senior US policy-makers have been wary about placing too many valuable intelligence assets on the one platform in case the platform malfunctioned or was destroyed during launch. Richelson, America’s Secret Eyes in Space (note 5) p.220.

21. Sweetman, ‘Spy Satellites: the Next Leap Forward’ (note 16) p.27. 22. A lenticular display shows images in alternating perspectives that allow the human brain to

fuse them into a three-dimensional picture. Mark Hewish, ‘Panning for Gold’, Jane’s International Defense Review, Dec. 2001, p.24.

23. In Kosovo, a US Predator became the first UAV to designate a target in combat. ‘Prospects for Unmanned Aerial Vehicles’, Strategic Comments 6/7 (Sept. 2000); John Tirpak, ‘Complications Overhead’, Air Force Magazine, April 1998, p.25.

24. Richelson, ‘High Flying Spies’ (note 7) p.52. 25. Stephen Grey, ‘Iron Fist Reaches out from the Other Side of the Globe’, The Australian, 19

Sept. 2001, p.1; John G. Roos, ‘Hunter-Killer UAVs’, Armed Forces Journal International, Jan. 2002, p.58.

26. These are the capabilities of the second generation Global Hawk II. ‘The Global Super Hawk’, Heads Up 230, 5 Dec. 2001, p.5.

27. John Tirpak, ‘The Robotic Air Force’, Air Force Magazine, Sept. 1997, <www.afa.org/magazine/1997robot.html>.

28. David A. Fulghum, ‘Computer Warfare Offense Takes Wing’, Aviation Week & Space Technology, 19 Jan. 1998, p.56.

29. Robert Wall, ‘Global Hawk In Australia Auditions For New Role’, Aviation Week & Space Technology, 30 April 2001, pp.32–3.

30. CIA Director, George J. Tenet, quoted in Robert K. Ackerman, ‘Intelligence at a Crossroads’, Signal, Oct. 2001, p.19. See also Mickey Galeotti, ‘Boom or Bust for the CIA’, Jane’s Intelligence Review, May 2002, p.51. Jeffrey Richelson canvasses the option of using clandestine operatives to conduct assassinations against especially egregious leaders and scientists involved in weapons programmes. Jeffrey T. Richelson, ‘When Kindness Fails: Assassination as a National Security Option’, International Journal of Intelligence and Counterintelligence 15/2 (Summer 2002) pp.243–79.

31. Artur Knoth, ‘March of the Insectoids’, Jane’s International Defence Review, Nov. 1994, pp.55–8.

36 TWENTY-FIRST CENTURY INTELLIGENCE

32. Pat Cooper, ‘Microbots Will Help U.S. Combat Bio-Chem Weapons’, Defense News, 11–17 Nov. 1996, p.22.

33. Cited in US Senate, ‘Foreign and Military Intelligence’, Final Report of the Select Committee to Study Governmental Operations with Respect to Intelligence Activities, 26 April 1976 (Washington, DC: US Government Printing Office 1976) p.351. Also known as the Church Report after its senior member, Senator Frank Church.

34. On this point, see Kovacs, ‘The Uses and Nonuses of Intelligence’(note 4) p.6; Richard Betts, ‘Analysis, War and Decision: Why Intelligence Failures Are Inevitable’, World Politics 31/1 (Oct. 1978) pp.61–89.

35. Hewish, ‘Panning for Gold’ (note 22) p.23; Barbara Starr, ‘CIA Looks to Web to Solve Data Overload’, Jane’s Defence Weekly, 23 July 1997, p.29.

36. This problem is not confined to the US and other states with advanced intelligence capabilities. South Korea’s Agency for National Security Planning has identified information management and the ability to organise and sort the massive raw material received from its agents as one of its highest priorities in the reform process initiated in early 1998 by its newly-installed head, Lee Jong Chan. Shim Jae Hoon, ‘Secrets and Lies: Reform of Spy Agency Aims to End its Political Meddling’, Far Eastern Economic Review, 9 April 1998, p.26.

37. Michael Handel (ed.), Intelligence and Military Operations (London: Frank Cass 1990) p.12. 38. For a useful discussion of the changing nature of intelligence analysis and the implications

of pull architecture, see Peter Sharfman, ‘Intelligence Analysis in an Age of Electronic Dissemination’, Intelligence and National Security 10/4 (Oct. 1995) pp.201–11.

39. Ibid., p.210. 40. US Senate, ‘Foreign and Military Intelligence’ (note 33) p.267; Michael Handel, War,

Strategy and Intelligence (London: Frank Cass 1989) pp.196 and 259. 41. A.D. McLennan, ‘National Intelligence Assessment: Australia’s Experience’, Intelligence

and National Security 10/4 (Oct. 1995) p.72. 42. Remarks made by Paul Wolfowitz in a short essay on intelligence which appears as a

commentary in Douglas MacEachin, ‘The Tradecraft of Analysis: Challenge and Change in the CIA’ (Washington DC: Working Group on Intelligence Reform Papers 1994).

43. This issue is addressed by a number of recent studies on intelligence, including Preparing for the 21st Century: An Appraisal of U.S. Intelligence, Commission on the Roles and Capabilities of the United States Intelligence Community (Washington, DC: U.S. Government Printing Office 1996), <www.access.gpo.gov/su docs/dpos/epubs/int/ index.html>. Otherwise known as the Brown Commission named after former US secretary of defense, Harold Brown.

44. On this point, see Gregory F. Treverton, ‘The Intelligence Agenda’, RAND Paper P-7941 (Santa Monica, CA: RAND 1995) p. 15.

45. Gregory F. Treverton, Reshaping National Intelligence for an Age of Information, (Cambridge: CUP 2001) pp.253–5.

46. These emerging threats, and their implications for security, are discussed in Alan Dupont, East Asia Imperilled: Transnational Challenges to Security (Cambridge: CUP 2001).

47. Barbara Starr, ‘Intelligence Community Coaxes its Analysts off the Fence’, Jane’s Defence Weekly, 29 Oct. 1994, p.20.

48. Mark Sommer, ‘Non-military Factors Bring New Meaning to Art of Spying’, The Jakarta Post, 15 Jan. 1996, p.5.

49. Strategic Assessment 1996: Instruments of U.S. Power (Washington DC: Institute for National Strategic Studies, National Defense Univ. 1996) p. 65.

50. Scott Pace, Kevin M. O’Connell and Beth E. Lachman, Using Intelligence Data for Environmental Needs: Balancing National Interests (Washington DC: Acquisition and Technology Policy Center, National Security Research Division, RAND 1997), <www.rand.org/publications/MR/MR799>.

51. Jim-Holden Rhodes and Peter A. Lupsha, ‘Gray Area Phenomena: New Threats and Policy Dilemmas’, Criminal Justice International 9/1 (Jan.–Feb. 1993) p.24.

37INTELLIGENCE FOR THE TWENTY-FIRST CENTURY

52. See, for example, Robert David Steele, ‘Open Source Intelligence: What Is It? Why Is It Important to the Military’, <www.oss.net/MILITARY/>; Strategic Assessment 1996: Instruments of U.S. Power (note 49) p.66.

53. Robert David Steele, ‘Private Enterprise Intelligence: Its Potential Contribution to National Security’, Intelligence and National Security 10/4 (Oct. 1995) p. 215.

54. Primarily due to the rapid reduction in the cost of microwave integrated circuits and digital processing. Starr, ‘USA Sees Vital Role for Commercial Satellites’ (note 6) p.21.

55. ‘High-resolution Commercial Satellite Imagery Available’, Jane’s International Defense Review, Feb. 2002, p.4.

56. Craig Covault, ‘Navy Enlists NASA In the War on Terror’, Aviation Week & Space Technology, 8 April 2002, p.31.

57. ‘Afghanistan in 3D’, Aviation Week & Space Technology, 7 Jan. 2002, p.17. On the other hand, the US was quick to reserve exclusive rights for commercial satellite imagery of Afghanistan useful for targeting and assessing battle damage. An additional benefit was that this material could be made available to coalition partners who did not have access to classified imagery. ‘US Renews Rights to Afghan Satellite Images’, The Times of India, 8 Nov. 2001, <www.timesofindia.com>.

58. On this point, see Bruce Berkowitz and Allan E. Goodman, ‘Why Spy – and How – in the 1990s?’, Orbis 36/2 (Spring 1992) p.270.

59. Preparing for the 21st Century: An Appraisal of U.S. Intelligence (note 43). 60. Hewish, ‘Panning for Gold’ (note 22) pp.25, 27. 61. ‘DIA Plans Virtual Intelligence Network’, Jane’s Defence Weekly, 2 Oct. 1996, p.11. 62. Steven Watkins, ‘USAF System Links Battlefield Intelligence Data’, Defense News, 12–18

Aug. 1996, p.12. 63. The four agencies in question were the CIA’s National Photographic Interpretation Center,

the imagery and dissemination functions of the Defense Intelligence Agency, the National Reconnaissance Office and the Defense Airborne Reconnaissance Office. Pat Cooper, ‘New U.S. Spy Agency Melds Mapping, Imaging Services’, Defense News, 14-20 Oct. 1996, p.68.

64. David A. Fulghum, ‘Planning Moves Deep Into Predictive Process’, Aviation Week & Space Technology, 18 June 2001, pp.175–6.

65. Gary Waters, ‘Operational Intelligence: Lessons >From The Gulf War’, in Anthony Bergin and Robert Hall (eds.), Intelligence and Australian National Security (Canberra: Australian Defence Studies Centre, Australian Defence Force Academy 1994) p.54; Berkowitz and Goodman, ‘Why Spy – and How – in the 1990s?’ (note 58) p. 273.

66. Barbara Starr, ‘US Rapid-targeting Data System Deployed to Gulf’, Jane’s Defence Weekly, 18 Feb. 1998, p.4.

67. Pat Cooper, ‘Air Force Stresses Big-Picture Reconnaissance’, Defense News, 12–18 Aug. 1996, p.14.

68. Preparing for the 21st Century (note 43). Some analysts do not categorise sensor-derived targeting information as intelligence but such reservations seem largely semantic and fail to recognise the integrated nature of modern C4ISR systems. See, e.g., Handel, (ed.), Intelligence and Military Operations (note 37) pp.11–12.

69. Pat Cooper, ‘Bosnia C2 System To Bridge Allied Communications Gap’, Defense News, 12–18 Aug. 1996, pp. 29 and 34.

70. ‘Defense Intelligence Carves a New Niche’, Signal, Oct. 2001, pp.21–2. 71. This concept is articulated in Freedman, The Revolution in Strategic Affairs (note 2) p.11. 72. Tim Weiner, ‘A Call for Complete Overhaul of CIA’, International Herald Tribune, 4 June

1998, p.3. Admiral David Jeremiah is a former vice chairman of the Joint Chiefs of Staff. 73. After the US commander of ‘Desert Storm’, General H. Norman Schwarzkopf. John Ferris

and Michael, I. Handel, ‘Clausewiz, Intelligence, Uncertainty and the Art of Command in Military Operations’, Intelligence and National Security 10/1 (Jan. 1995) p.50.

74. Bruce Berkowitz, ‘Information Technology and Intelligence Reform’, Orbis 41/1 (Winter 1997) p.114.

75. Ball, Desperately Seeking bin Laden (note 5) p.63.

38 TWENTY-FIRST CENTURY INTELLIGENCE

76. Mark Gaylord, ‘Money Laundering in Asia’, in Ann Lodl and Zhang Longguan (eds.), Enterprise Crime: Asian and Global Perspectives (Chicago, IL: Office of Int. Criminal Justice with Shanghai Bureau of Justice and The East China Inst. of Politics and Law, Univ. of Illinois 1992) pp.82–3.

77. Bandwidth problems are being addressed and there is growing optimism that they will not be a serious constraint on the development of more capable intelligence architectures. Bandwidth capacity is doubling every four months compared with 18 months for computer power. ‘Defense Intelligence Carves a New Niche’, Signal, Oct. 2001, p.21.

78. William J. Broad, ‘New Technology Pares Spies’ Edge’, International Herald Tribune, 21 Sept. 2001, p.15.

79. Information warfare is defined here ‘as activities undertaken by governments, groups or individuals to gain electronic access to information systems in other countries either for the purpose of obtaining the data in such systems, manipulating or fabricating the data, or perhaps even bringing the systems down, as well as activities undertaken to protect against such activities’. Preparing for the 21st Century (note 43).

80. Defense News, 14–20 Oct. 1996, p.18. 81. This is certainly the belief of the former head of the US National Security Agency, Lt. Gen.

Ken Minihan. Pat Cooper, ‘DoD Directive Links Info War Intelligence, Operations’, Defense News, 14–20 Oct. 1996, p.84.

82. Bradley Graham, ‘In Cyberwar, A Quandary Over Rules and Strategy’, International Herald Tribune, 9 July 1998, <www.int.com>. See also David A. Fulghum, ‘Cyberwar Plans Trigger Intelligence Controversy’, Aviation Week & Space Technology, 19 Jan. 1998, pp.52–3; David A. Fulghum, ‘Computer Combat Rules Frustrate the Pentagon’, Aviation Week & Space Technology, 15 Sept. 1997, pp. 67–8.

83. The Global Information Network is defined as ‘a world-wide network of information sources, archives, consumers and architectures’. Freedman, The Revolution in Strategic Affairs (note 2) p.49.

84. Robert Steele quoted in Leonard Davis, ‘Officials Question Sat Ability To Track’, Defense News, 17–23 Sept. 2001, p.6.

39INTELLIGENCE FOR THE TWENTY-FIRST CENTURY