Understanding Intelligence Cycle (IC)

In this paper I will be assessing to what extent the Intelligence Cycle (IC) can be used to understand the creation of a finished intelligence product. I will discuss the individual parts of the IC in depth, and derive the utility and theoretical foundations of the cycle. A discussion of the advantages and disadvantages of the IC will act as a tool of understanding the creation of finished intelligence. This will prove beneficial when comparing it to other processes and/or models that have been developed as a way to understanding the intelligence system.

Intelligence can be traced back as far back as history will allow, and Sun Tzu comments, “It is only the enlightened ruler and the wise general who will use the highest intelligence of the army for the purposes of spying, and thereby they achieve great results.”[1] Intelligence is an abstract concept, and Lacquer comments that “…all attempts to develop ambitious theories of intelligence have failed.”[2] However, the best definition of intelligence incorporates the process of intelligence collection and the means to which it is used. Gill and Phythian define intelligence as, “an umbrella term referring to the range of activities – from planning and information collection, analysis and dissemination – conducted in secret; aimed at providing forewarning of threats and for the implementation of policy up to and including covert action.”[3] Finally, the Dictionary of Military and Associated Terms provides consensus, defining intelligence as: “Information and knowledge about an adversary obtained through observation, investigation, analysis, or understanding.”[4]

In order to understand intelligence we must first understand the means by which intelligence is collected. The Intelligence Cycle is taught as a scientific process to rationalize and normalize the process of gathering intelligence. It provides a simplified picture of the relationship between ‘Intelligent Consumers’ – which are at the beginning and end of each cycle and ‘Intelligent Producers’ – which account for the rest of the cycle. This is also echoed in the private and business sectors. The main elements to the Intelligence Cycle are; Requirements, Planning/Direction, Collection, Processing, Analysis, and finally Reporting/Dissemination. This can be seen in Figure 1 (appendix) which represents the Intelligence Cycle graphically.

The first part of the cycle is requirements. The intelligence consumer (also the intelligence end user) submits an intelligence requirement/request. For example, the consumer could range from a policy official needing to know the intention-probability threat level of a suspected terrorist attack, to a public health body that needs to assess the validity of its response to a mass casualty producing event such as a chemical/biological weapon. At this point it is important to distinguish the UK requirements-driven IC and the USA data-driven [traditional] IC due to the resource disparities between the nations. However, a broad ‘requirement’ such as “potential terror threats to the nation” for all intents and purposes will be the same in data and requirements driven Intelligence Cycles.

Once the intelligence requirements are established then the planning function of the IC can start to ascertain how this requirement will be fulfilled. This is essentially the management and co-ordination of the intelligence process – from identifying the need for information, to delivery of an intelligence product to an end user. The “Five W’s” (Who, What, When, Where and Why as well as How) are the start points for identification of intelligence needs into requirements. These points define the framework for decision makers who establish the ‘Essential Elements of Information'[5].

At this point in the cycle the requirements have been established, and some form of direction has been provided to the intelligence producers. They can now embark on collecting the information required to fulfil the intelligence requirement. This can be through a variety of ways; the easiest and most common route is through ‘open source’ intelligence (OSINT). This includes the media, newspapers, academic literature, foreign and domestic broadcasts etc. In fact, open source reporting is integral to the CIA’s analytical capabilities. Whilst an intelligence analyst can access most OSINT with ease, there is a divide between agencies with the responsibilities and abilities of collection. For example, National Security Agency (NSA) for Signal Intelligence (SIGINT), the Central Intelligence Agency (CIA) for human intelligence (HUMINT) and the National Geo-Spatial Intelligence Agency (NGA) for Imagery Intelligence (IMINT). It is often noted as a paradox that 95% of intelligence comes from open source information, yet 95% of the intelligence budget is spent on obtaining information from covert actions and clandestine missions.

Collection is the prelude to the most intensive part of the IC – the processing and exploitation of the information collected. As one former National Security Agency director said, “Gleaning hard facts from the avalanche of information was like trying to take a drink of water from a fire hose.”[6] Therefore it is important to covert the masses of information into a form that can be utilized by intelligence analysts. Exploitation “includes integrating, evaluating and analyzing all available data – which is often fragmentary and even contradictory”.[7] It is also important to note that “raw intelligence is often referred to as ‘the dots’ – individual pieces of information. Finished intelligence reports the ‘connected dots’.”[8] A critical note of the 9/11 intelligence failures was that it was not due to a lack of information/intelligence. In fact, analysts were drowning in an ocean of information, but they were unable to ‘connect the dots’ to provide a comprehensive picture. This further compounds the idea that processing and exploitation of information is crucial to the intelligence cycle.

The final part of the IC is dissemination/reporting. This is the distribution of the finished intelligence to the intelligence consumer who initiated the request. This can be in the way of brief one page reports, lengthy case studies or a presentation. The intelligence customers then make decisions – operational, strategic and policy – based on the information received. This may in turn lead to more intelligence requirements levied on the intelligence producers, which restarts the cycle again.

We have determined that the IC is a structural model to help us understand the creation of an intelligence product, but this paper now looks to what extent this model is valid, and empirical evidence where it seems the IC has broken down. There are instances where it seems the IC model has been a success; “For instance, the Bush administration based its estimate of North Korea’s nuclear aspirations largely on satellite imagery depicting activity at the Yongbyon nuclear facility and air samples suggesting that spent nuclear fuel was being reprocessed into weapons-grade plutonium. It concluded that Kim Jong-Il’s government was on the verge of developing capabilities menacing enough to warrant a forceful U.S. response.”[9]

However, it is said that whilst the IC is simple in theory, it is messy in practise. The simplicity of the model explained above does not take into consideration factors that may affect the overall outcome. There are conceptual weaknesses with components of the IC, and this affects the integrity of the model as a whole.

Firstly, Gill and Phythian comment that the IC does not capture the dynamic nature of intelligence; rather the IC is seen as a static model. Whilst neat in theory, in times of emergencies and national crisis’ the IC is bypassed or it may even be enveloped into one user. Semerson comments that “Intelligence analysts dread having their work overtaken by events, for fear of being cut out of the decision-making loop in the future”[10] and Berkowitz adds value commenting that, “…Intelligence planners short-circuited the formal process when high-level consumers needed information in a hurry and had the clout to have their way.”[11] This increases the risk of the IC being played out in one individual, and it raises questions of accountability and lack of objectivity by an analyst. Furthermore, the risk of bias within the individual in the rush to provide timely intelligence may infer bias onto the policy directive. Moreover, the IC does not take account of counter intelligence/covert action or false intelligence. The impact of false intelligence could disrupt the validity of the IC, and again may infer bias onto the policy directive.

Another major problem with the IC is the increased use of ‘stove piping’. In the traditional sense, ‘stove piping’ keeps the output of different collection systems separated from one another. The analogy is that of a chimney – that smoke is directed through a small pipe where it is needed, and does not fill the room in general. However, this has various derogatory implications in regards to the IC. Firstly, ‘stove piping’ prevents one discipline from cross checking another. This prevents “…each unit [in the Intelligence Cycle] letting the unit next door have a place in the ‘chop chain’, allowing it the opportunity to add its expertise to the final product.”[12] Moreover, ‘stove piping’ bypasses the regular analysis of raw information and sends only raw intelligence to the Intelligence Consumer that supports a particular position/policy they have taken.

This leads on to another problem associated with the IC – that the classic notion of intelligence used to inform and drive policy has been usurped by intelligence that is channelled to support a policy stance. This ‘politicization’ of intelligence is best demonstrated by the use of intelligence in the Second Gulf War (2003 – present day). Intelligence has become a support for policy decisions already made, not to drive a policy decision. It is argued that the United States had already committed itself to the war in Iraq, and was searching for intelligence to indicate evidence of WMDs in Iraq.

The arguments below compound the idea that the IC is ‘messy in practice’. Whilst the United States is an affluent nation with vast resources to dedicate to the intelligence community, it is not omniscient. Semerson comments that, “Unless top political leaders rigorously think out their priorities and summon the will to impose discipline on the process… the Intelligence Cycle can degenerate into an effort to learn everything about everything, everywhere in the world – a fool’s errand.”[13] This argument is further compounded by the volumes of raw data that is compiled (by the US in particular) and then “shelved until the administration deems the subject matter a priority”[14]

Another problem, simple as it may be, is that it is impossible to know everything, and it is particularly difficult to penetrate closed societies – especially those obsessed with internal security such as Iraq and North Korea. Therefore, the context of the IC is inapplicable to these societies. Finally, Semerson comments that, “Bits of key information routinely slip through the cracks, detracting from the overall intelligence estimate.”[15] This is further compounded by critics who claim, “It is never a satisfactory excuse to say ‘if we just had more information’. Analysts already have more information than they can digest.”[16]

At this point in the paper we can now introduce other processes/models to help understand the intelligence process. We start by distinguishing between the ‘Traditional’ IC and Robert Clark’s ‘Target-Centric’ approach. Berkowitz comments that, “The [traditional] intelligence cycle reflects the best thinking of how an information service should work from the late 1940s and 1950s.”[17] The problem with the traditional IC is that each stage is isolated where each part has a designated task. There is limited scope for feedback or to for intelligence consumers to ask questions. Kent argues that there should be a collective responsibility for judgement, as opposed to the traditional IC which, “separates collectors, processors and analysts and too often results in ‘throwing information over the wall’ to become the next persons responsibility. Everybody neatly avoids responsibility for the quality of the final product and it is more predictable and therefore more vulnerable to an opponent’s countermeasures.”[18] The target-centric approach attempts to refine the intelligence process in such a way that it is a network rather than a cycle. Sherman Kent adds consensus by supporting a network approach to intelligence where analysts are directly accountable for the work. Finally, Hulnick compounds this view stating that “collection and analysis, which are supposed to work in tandem, in fact work more properly in parallel.”[19]

Both Jardines and Clark argue that, “a more target-centric, iterative and collaborative approach…would be far more effective than the current traditional intelligence cycle.”[20] Clark summarizes, “With a target-centric approach to intelligence analysis, intelligence is collaborative, because this model creates a system where it can include all contributors, participants, and consumers. Each individual can question the model and get answers along the way. The target-centric model is a network process where the information flows unconstrained among all participants, who also focus on the objective to create a shared picture of the target”[21] Finally, Berkowitz adds consensus claiming that, “the best way to ensure that an intelligence product meets the needs of the consumer is to put the consumer in touch with the analysts preparing the product.”[22] This would have the effect of, “minimizing the distance between intelligence producers and intelligence consumers, permitting enough interaction so that the product is naturally tailored to the needs of the consumer.”[23]

However, a Target-Centric model, by its nature is a network process that is more time-consuming than the traditional IC, and therefore the finished product is likely to take a longer time to reach the intelligence consumer. This may explain why the IC is preferred above other models to explain the finished intelligence product.

Another process is John Boyd’s ‘OODA Loop’ (see Figure 2). OODA (Observe, Orientate, Decision, and Action) is more action orientated and resembles a spiral rather than a continuous cycle. It was developed originally for air-to-air fighter combat decision-making/action. After the action, the subject observes again, to see the effects of the action. If the cycle works properly and the subject has initiative, then he can orient, decide, and act even faster in the subsequent iterations of the OODA Loop. This model differs from the IC and Target-Centric because OODA focuses on a single decision maker, and Greene comments that, “The proper mindset is to let go a little, to allow some of the chaos to become part of his [decision makers] mental system, and to use it to his advantage by simply creating more chaos and confusion for the opponent. He funnels the inevitable chaos of the battlefield in the direction of the enemy.”[24] Eventually, if the process works as intended, the subject will ‘get inside the opponent’s loop’. When the subject OODA Loop dominates the opponent’s, the subject is acting repeatedly, based on rational choices, while the opponent is still trying to understand what is happening. It is important to note that the OODA Loop is less applicable in comparison to the IC, and it tends to be discussed not as a stand-alone model, but in the context of the IC.

Lionel Giles, ‘The Art of War by Sun Tzu – Special Edition.’ (El Paso Norte Press, 2005)
Walter Laqueur, ‘A World of Secrets: The Uses and Limits of Intelligence’ (New York: Basic Books, 1985) p8
Peter Gill, Mark Phythian, ‘Intelligence in an insecure world’ (Polity Press, 2000) p2.
Joint Chiefs of Staff, ‘Department of Defence Dictionary of Military and Associated Terms’ (Joint Publication 1-02, 2001) p208.
‘The Intelligence Cycle’ http://www.tech-writer.net/intelligencecycle.html (accessed 15th November2009)
Bruce Berkowitz, ‘Best truth: Intelligence in the information age.’ (Yale University Press, 2002) p66
‘The Intelligence Cycle’ http://fas.org/irp/cia/product/facttell/intcycle.htm (Accessed 15th November 2009)
‘The Intelligence Cycle’ http://www.fbi.gov/intelligence/di_cycle.htm (Accessed 15th November 2009)
Semerson, ‘The Limits of Gathering and Interpreting Intelligence’, (ESAI Weekly Intelligence Briefing, 28/07/2003) p2
Semerson, ‘The Limits of Gathering and Interpreting Intelligence’, (ESAI Weekly Intelligence Briefing, 28/07/2003) p2
Bruce Berkowitz, ‘Best truth: Intelligence in the information age.’ (Yale University Press, 2002) p73
Bruce Berkowitz, ‘Best truth: Intelligence in the information age.’ (Yale University Press, 2002) p73
Semerson, ‘The Limits of Gathering and Interpreting Intelligence’, (ESAI Weekly Intelligence Briefing, 28/07/2003, p1)
Semerson, ‘The Limits of Gathering and Interpreting Intelligence’, (ESAI Weekly Intelligence Briefing, 28/07/2003, p1)
Semerson, ‘The Limits of Gathering and Interpreting Intelligence’, (ESAI Weekly Intelligence Briefing, 28/07/2003, p3)
‘Intelligence Analysis and Cycle’ http://www.intelligencesearch.com/ia084.html (accessed 17th November 2009)
Bruce Berkowitz, ‘Best truth: Intelligence in the information age.’ (Yale University Press, 2002) p69
Eliot Jardines, ‘Testimony of Eliot A. Jardines’, in ‘Using Open Source Effectively: Hearings before the Subcommittee on Intelligence, Information and Terrorism Risk Assessment, Committee on Homeland Security’ (21 June 2005), p14
Arthur Hulnick, “What’s Wrong with the Intelligence Cycle.” Intelligence and National Security 21, no. 6 (Dec. 2006) p961
Eliot Jardines, ‘Testimony of Eliot A. Jardines’, in ‘Using Open Source Effectively: Hearings before the Subcommittee on Intelligence, Information and Terrorism Risk Assessment, Committee on Homeland Security’ (21 June 2005), p3
Robert Clark, ‘Intelligence Analysis: A Target-Centric Approach’ (CQ Press, 2006), p14
Bruce Berkowitz, ‘Best truth: Intelligence in the information age.’ (Yale University Press, 2002) p74
Bruce Berkowitz, ‘Best truth: Intelligence in the information age.’ (Yale University Press, 2002) p74
Robert Greene, ‘OODA and You’, http://www.powerseductionandwar.com/archives/ooda_and_you.phtml, (Accessed 22nd November 2009)