Portal de Periódicos da CAPES

An Army of One: Assessing CBRN Pursuit and Use by Lone Wolves and Autonomous Cells

2013; Taylor & Francis; Volume: 26; Issue: 1 Linguagem: Inglês

10.1080/09546553.2014.849945

1556-1836

Gary Ackerman, Lauren E. Pinson,

Terrorism, Counterterrorism, and Political Violence

Abstract In recent years, much of the literature on terrorist pursuits and use of unconventional weapons has focused on al-Qaeda and other Islamic extremist groups. Some scholars posit lone actors may differ from organizations when it comes to the observation that larger organizations are more likely to engage in chemical, biological, radiological, and nuclear (CBRN) activity, but leave this as an open question because of a prior lack of data on lone actors. Furthermore, new technologies and expanding networks may also spawn more deleterious lone actors, including what have been referred to as "super-empowered individuals." This article utilizes the Profiles of Incidents Involving CBRN by Non-State Actors Dataset and the Radiological and Nuclear Non-State Actors Database to illustrate the characteristics of CBRN pursuit by lone wolves and autonomous cells over time. Initial findings confirm the general perceptions that, historically, lone actors have engaged in cruder, smaller scale, and less frequent CBRN plots and attacks than their formal terrorist organization counterparts, but suggest that such actors do present a genuine threat, at least in terms of the potential for mass disruption. Additionally, while they share several similarities with more formal organizations, lone actors and autonomous cells tend to be motivated less by collective religious or ethno-nationalist concerns than larger organizations, focusing more on narrow or solipsistic drivers. Lone actors and autonomous cells also appear to largely calibrate their CBRN ambitions to their capabilities and their CBRN actions tend to receive better media coverage than formal terrorist organizations. Keywords: biologicalCBRNchemicallone actornuclearradiologicalterrorismweapons of mass destructionWMD Acknowledgments The authors wish to thank the Defense Threat Reduction Agency and the Department of Homeland Security for their support in the development of the databases described in this article. They also thank the editors of this journal issue for their helpful comments. Notes a There are nine cases that the coders of RANNSAD marked as potentially apocryphal. All analyses mentioned below were performed both including and excluding these potentially apocryphal cases and the text will note where this affects the results. The four models that were used in all of the analyses were: a) the model that included the 26 likely lone actors/autonomous cells and all cases (the base model); b) the model that included only the 19 confirmed lone actors/autonomous cells and all cases; c) the same as (a) but with apocryphal cases removed; and d) the same as (b) but with apocryphal cases removed. b In certain cases, for instance where a perpetrator was interdicted in possession of raw uranium, it is unclear whether the intent was to use the material to construct a nuclear or a radiological weapon. Roger Griffin, "From Slime Mould to Rhizome: An Introduction to the Groupuscular Right," Patterns of Prejudice 37, no. 1 (March 2003): 27–50. We define "weapons of mass destruction" in this context as chemical, biological, radiological, and nuclear (CBRN) weapons that, if used, would inflict catastrophic casualties, widespread social disruption, or devastating economic consequences greater than those brought about by the largest conventional attacks carried out thus far. For a discussion of this definition and the merits of this term, see Gary Ackerman and Jeremy Tamsett, Introduction to Jihadists and Weapons of Mass Destruction (Boca Raton, FL: CRC Press, 2009), xix–xxii. The field of synthetic biology is concerned with designing and building new biological materials (e.g., enzymes, proteins, lipids not already found in nature), devices, systems, and/or functions, especially by (re)programming DNA circuits to manipulate cellular behavior. Nanoscale engineering refers to designing, manipulating, and building nanoscale (objects measuring up to thousands of nanometers, which are one-billionth of a meter) materials, structures, and systems, and integrating these devices with other systems. A chemical microreactor is a small-scale device (e.g., a square centimeter) in which chemical reactions take place using tiny channels and which can provide greater control over the temperature of the reactions and allow for continuous flows of chemicals, in comparison with traditional large and often inefficient "batch" reactors. Even the efficacy of traditional asymmetric methods such as poisoning and contamination (common since antediluvian times) required luck and precise targeting and was usually only limited to a single water or food source. Gary Ackerman, "Nuclear Terrorism: Assessing the Threat to the Homeland" Testimony before the United States Senate Committee on Homeland Security and Governmental Affairs (Washington, D.C., April 2, 2008). John Robb, Brave New War: The Next Stage of Terrorism and the End of Globalization (Hoboken, NJ: John Wiley Sons, 2007). In this regard, artificial intelligence theorist Eliezer Yudkowsky has only semi-jokingly stated that "the IQ level required for a single individual to destroy the world decreases by one point every year." Oxford University Global Risks Conference (Oxford, United Kingdom, July 19, 2008). Thomas Friedman, one of the first commentators to engage with this phenomenon, in fact suggests that the base driving force behind the rise of these "super-empowered individuals" is globalization. The faster, cheaper, and deeper access that people have to "corporations, countries and individuals" has both radicalized and empowered some individuals in Western society. Thomas L. Friedman, Longitudes and Attitudes: Exploring the World After September 11 (New York: Farrar, Straus & Giroux, 2002); Adam Elkus, "Night of the Lone Wolves," Defense and the National Interest Special, November 29, 2007, http://www.dnipogo.org/fcs/elkus_lone_wolves.htm. As such, they are neither inherently good nor evil, but can be normatively defined by their actions. Elkus, "Night of the Lone Wolves" (see note 8 above). In the case of chemical weapons, for example, Smithson argues that "Whereas terrorist groups tend to be expedient and go for proven, readily acquired weapons, lone actors may relish the technical challenge of making a novel nerve agent" (79). Amy Smithson, "Indicators of Chemical Terrorism," in Magnus Ranstorp and Magnus Normark, eds., Unconventional Weapons and International Terrorism: Threat Convergence in the Twenty-First Century (New York: Routledge, 2007), 67–95. John Parachini, "Comparing Motives and Outcomes of Mass Casualty Terrorism Involving Conventional and Unconventional Weapons," Studies in Conflict and Terrorism 24, no. 5 (September–October 2001): 389–406. See Sammy Salama and Lydia Hansell, "Does Intent Equal Capability? Al-Qaeda and Weapons of Mass Destruction," The Nonproliferation Review 12, no. 3 (2005): 615–653; also John Mueller, "Radioactive Hype," The National Interest 90 (September–October 2007): 59–65. Adam Dolnik, "13 Years Since Tokyo: Revisiting the 'Superterrorism' Debate," Perspectives on Terrorism 2, no. 2 (2008): 3–11, 10. In this regard, Kenney makes the important distinction between techne (technical knowledge that underpins specific applications) and mētis, which includes such factors as local knowledge, practical skills, and forms of cunning. Michael Kenney, "'Dumb' Yet Deadly: Local Knowledge and Poor Tradecraft Among Islamist Militants in Britain and Spain," Studies in Conflict and Terrorism 33, no. 10 (2010): 912 and passim. While some people have questioned whether Dr. Ivins was the true perpetrator of the anthrax letters or whether he did in fact operate alone based on certain inconsistencies in the official reporting of the case, no one has presented sufficient evidence to seriously undermine the U.S. government's version of events, which asserts that Dr. Ivins was the sole person responsible for the formulation, production, and dissemination of the deadly anthrax spores. All of these examples are drawn from the POICN Database (see the following section for a full description). National Consortium for the Study of Terrorism and Responses to Terrorism (START), Profiles of Incidents involving CBRN by Non-state Actors Database [data file] (College Park, MD: Author, 2012). Charles P. Blair, Kelsey Gregg, and Jonathan Garbo, Norway's Anders Breivik: Weapons of Mass Destruction and the Politics of Cultural Despair (Washington, DC: Federation of American Scientists 2011), http://www.fas.org/blog/ssp/2011/07/norways-anders-breivik-weapons-of-mass-destruction-and-politics-of-cultural-despair.php#. "Al-Qaeda Magazine Urges Chemical, Biological Strikes Against Foes," Global Security Newswire, May 3, 2012. In contradistinction to Spaaij (op. cit.), who asserts that there has only been a single case of threatened CBRN terrorism by a lone actor. Ramon Spaaij, "The Enigma of Lone Wolf Terrorism: An Assessment," Studies in Conflict and Terrorism 33, no. 9 (2010): 854–870. Indeed, as will be discussed below, the POICN Database contains at least 18 CBRN terrorist attacks by lone actors, let alone threats. Especially when dominated by a few high-profile events, such as the "anthrax letters" or the 1995 sarin attacks on the Tokyo subway system by the Japanese Aum Shinrikyo cult, the potential for selection bias is high. Jeffrey D. Simon, "Lone Operators and Weapons of Mass Destruction," in Brad Roberts, ed., Hype or Reality: The 'New Terrorism' and Mass Casualty Attacks (Alexandria, VA: Chemical and Biological Arms Control Institute, 2002), 69–81. George Michael has an entire chapter in his book on Lone Wolf Terror and the Rise of Leaderless Resistance devoted to WMD, but this chapter is mostly a general survey of CBRN terrorism, not fully defining distinguishing characteristics between lone actors/autonomous cells and large terrorist groups in this regard. See George Michael, "Weapons of Mass Destruction and Leaderless Resistance," in Lone Wolf Terror and the Rise of Leaderless Resistance (Nashville, TN: Vanderbilt University Press, 2012), 101–118. See, for example, Jonathan B. Tucker and Amy Sands, "An Unlikely Threat," Bulletin of the Atomic Scientists 55, no. 4 (1999): 46–52; Jeffrey M. Bale and Gary A. Ackerman, "Profiling the WMD Terrorist Threat," in Stephen M. Maurer, ed., WMD Terrorism: Science and Policy Choices (Cambridge, MA: MIT Press, 2009), 11–46; Dolnik, "13 Years Since Tokyo" (see note 14 above); and Bob Graham, Jim Talent, Graham Allison, Robin Cleveland, Steve Rademaker, Tim Roemer, et al., World at Risk: The Report of the Commission on the Prevention of WMD Proliferation and Terrorism (New York: Vintage Books, 2008). Examples include TEVUS, BAAD, and TOPs: National Consortium for the Study of Terrorism and Responses to Terrorism (START), Terrorism and Extremist Violence in the United States (TEVUS) Dataset [data] (College Park, MD: Author, 2012); the Big Allied and Dangerous (BAAD) Dataset described in Victor H. Asal and R. Karl Rethemeyer, "The Nature of the Beast: Terrorist Organizational Characteristics and Organizational Lethality," Journal of Politics 70, no. 2 (2008): 437–449; and Memorial Institute for the Prevention of Terrorism (MIPT), Terrorist Organization Profiles (TOPs) [data] (Oklahama City: Author, 2008), http://www.start.umd.edu/start/data_collections/tops/. Victor Asal, Gary Ackerman, and R. Karl Rethemeyer, "Connections Can Be Toxic: Terrorist Organizational Factors and the Pursuit of CBRN Weapons," Studies in Terrorism and Conflict 35 (2012): 229–254. Victor Asal and R. Karl Rethemeyer, "Islamist Use and Pursuit of CBRN Terrorism," in Gary Ackerman and Jeremy Tamsett, eds., Jihadists and Weapons of Mass Destruction (Boca Raton, FL: CRC Press, 2009): 335–358. Asal, Ackerman, and Rethemeyer, "Connections Can Be Toxic" (see note 25 above). In some cases, multiple perpetrators are connected to a single event, which results in the total number of (non-unique) perpetrator records being greater than the total number of events. However, no lone actors or autonomous cells are recorded as being involved in events with multiple perpetrators. Unnamed cell applies when an event is perpetrated by an unnamed, unaffiliated cell with fewer than 10 people. Assessing the relationship between actor type and event type yields a Gamma of 0.174 and a Somers' d (treating event type as dependent) of 0.128 at a 2.2% level of significance, which indicates a modest correlation between the two variables. The peak of CBRN use by formal organizations in 1998 represents poisonings perpetrated by the East Turkistan Liberation Organization. An event coded as heightened government interest must fulfill at least one of the following three categories: the event resulted in at least five total casualties; the event involved a CBRN agent that is classified as a warfare agent; or the event involved either the use of or a plot to create a weaponization of the agent in at least a moderately sophisticated manner. When analyzing the relationship between actor type (with all probable and possible lone actors/autonomous cells included) and the heightened interest variable, the Gamma value is −0.314 at a significance level of 0.3%, while the Spearman correlation is fairly weak at −0.143. Sophistication captures the overall level of complexity required in successfully acquiring, producing, and delivering the harm agent. When analyzing the relationship between actor type and attack sophistication, one obtains a Gamma of −0.581 at a significance level of <.001 and a Somers' d value (treating attack sophistication as the dependent) of −0.373. Analyzing the relationship between actor type and motivation yields a Cramer's V of 0.668 at a level of significance of <.001, indicating a fairly strong relationship. High source validity is equal to three or more independent sources or two independent sources with competing biases used for coding data. There are several general examples of perpetrator-based databases in the terrorism and political violence field, for example Big Allied and Dangerous (BAAD) Dataset and United States Extremist Crime Database (ECDB). POICN has a component that maps longitudinal data on perpetrators, but so far this is limited to established organizations. Gary A. Ackerman, Charles P. Blair, and Maranda Sorrells, Radiological and Nuclear Non-State Adversaries Database (RANNSAD) (College Park, MD: National Consortium for the Study of Terrorism and Responses to Terrorism, 2011). The database can be accessed from http://start.umd.edu/start/announcements/announcement.asp?id=396 or the Dataverse Network at http://dvn.iq.harvard.edu/dvn/dv/start/faces/study/StudyPage.xhtml?globalId=hdl:1902.1/16258&studyListingIndex=0_e274926257afe9192bedd589b32b. A companion database, the Chemical and Biological Non-State Adversaries Database (CABNSAD) is currently under development and is scheduled for completion in 2013. RANNSAD contains at least some information on 63 lone actors or members of groups. In addition to the relatively small sample size and missing data, one must remain mindful of the potential for selection bias. For example, although the majority of unsuccessful plots appear to have been interdicted (19) rather than abandoned (1), this may simply be because the former activity is more likely to appear in open-source reports. The database as it is available lists 14 perpetrators as being of unknown type (with 19 recognized lone actors/autonomous cells and 12 formal organizations making up the total), but the authors, by closely examining the qualitative descriptions of each case, were able to classify 11 of these unknowns into likely categories of actor. All statistical analyses, however, were performed on both sets—the more conservative one consisting of 31 perpetrators of known type and the more extensive one consisting of 42 perpetrators of likely type. We will note wherever the substantive results differ. Performing tests of association on the data (excluding indeterminate cases) yields a significant difference between the two different sets of actors, with Cramer's V = 0.43 and significant at the 5% level in the base model, and similar results in the other models (although the model using the likely categorization and excluding the potentially apocryphal events is only significant at the 7.1% level). However, these results are somewhat suspect since the Chi-square test yields 5 of 8 cells with an expected value of less than 5. For example, the cases listed under the "Nuclear Only" and "Radiological and Nuclear" category in Table 4 include a break-in at a South African nuclear facility and an attempt to smuggle enriched uranium in France that might have had a connection to a terrorist group (see RANNSAD cases 25 and 32). The relationship between the variables in the base model is significant at the 0.4% level, with Cramer's V = 0.828, which indicates a strong correlation. The other models yield similar results. All individuals or group members in the database are male. Since the Highest Activity Type is an ordinal variable (after removing unknowns from consideration), the Gamma and Somer's d tests were indicated. Treating the Highest Activity Type as the dependent variable, the Somer's d value for the base model was 0.53 at a significance level of <1%, while the symmetric Gamma measure was 0.64. The other three models showed similar results. One potential cause for concern is that the limited number of incidents led to the Chi-square test, while strongly significant, having 70% of its cells with an expected count of less than 5. Indeed, when a perpetrator pursued a radiological weapon, he almost always proceeded to at least the acquisition of material stage and often to the use stage as well. Measures of association are equivocal in this case, with the more conservative categorization (models (b) and (d)) having significance at the 5% level for certain measures (e.g., Spearman correlation), but not others. The other models show far lower levels of significance. In May 2002, Gu Jiming, a nuclear medical researcher at a hospital in Guangdong Province, China hid a container of radioactive Iridium-192 pellets above the ceiling panels of the office of a business rival in order to sicken him through radiation poisoning. A total of 75 people at the hospital suffered radiation sickness (RANNSAD Case 17), making this the largest recorded instance of intentional radiological injuries caused by a non-state actor. Although not meeting the common threshold of 100 casualties for a mass-casualty incident, it does show that large numbers of casualties are possible (although few of these will ever end up being non-stochastic radiological fatalities). The remaining 12 use cases in RANNSAD caused a total of 5 casualties. None of the four models was significant, even at the 10% level. None of the four models was significant, even at the 10% level. The degree of association is very strong, with Cramer's V = 0.91 and significant to <0.1% in the base model, and similar results in the other three models. This aspect has received far less attention in the literature on RN terrorism. For perpetrators consisting of multiple individuals, if age data was available for more than one member, the average for the individual members was used. For the base model, Somer's d is −0.824 and significant at less than 0.1%, while the Gamma is −1 at a similar level of significance. The other models yield substantially similar results. None of the four models was significant, even at the 10% level, when tested using the Somer's d and Gamma tests, which are appropriate for ordinal data. Walter Laqueur, The New Terrorism: Fanaticism and the Arms of Mass Destruction (Oxford: Oxford University Press 1999), 269. Spaaij, "The Enigma of Lone Wolf Terrorism" (see note 20 above) declares in 2010 that "there is no empirical evidence to support Laqueur's claim that lone individuals are among the most likely candidates to use weapons of mass destruction." POICN; includes both confirmed and probable LA/AC actual attacks. It should be noted that this does not mean that there are in fact no differences in these variables, but that with the data currently available to us, there was no statistically meaningful difference. One of the observations, that, post-event, LA/AC RN perpetrators are more likely to be apprehended than members of traditional terrorist organizations (perhaps because of more meager logistical resources like safe havens and forged documentation) is interesting, but not particularly relevant to preventing the attacks in the first place, except of course with respect to preventing follow-on attacks. Level of education was not significant in our small sample. However, this might prove to be a key distinguishing variable and it is recommended that future research return to investigate it when a larger sample becomes available. After all, as the Commission on the Prevention of WMD Proliferation and Terrorism report states in the context of bioterrorism, but could apply equally to other weapon types: "the United States should be less concerned that terrorists will become biologists and far more concerned that biologists will become terrorists." Graham et al., World at Risk (see note 23 above), 11. David Hume, An Enquiry Concerning Human Understanding, eds. L. A. Selby-Bigge and P. H. Nidditch, 3rd. ed. (Oxford: Clarendon Press 1777 (1975)). These are terms often used by futurists to describe high-impact events with small, yet unknown, probabilities of occurrence, often representing a discontinuous leap from prior experience. See Nicholas Taleb, The Black Swan: How the Improbable Rules the World and Why We Don't Know It (New York: Random House, 2007); John Petersen, Out of the Blue: How to Anticipate Big Future Surprise (Lanham, MD: Madison Books, 1999). Additional informationNotes on contributorsGary A. Ackerman Gary A. Ackerman is the Director of the Special Projects Division at the National Consortium for the Study of Terrorism and Responses to Terrorism (START), University of Maryland. Lauren E. Pinson Lauren E. Pinson is a Senior Research/Project Manager at START and a PhD student at Yale University.