Two high-profile attacks by terrorists on highly secure military bases in Pakistan, the first on the General Headquarters of the Pakistan Army in Rawalpindi in October 2009 and the second on the naval aviation base at PNS Mehran near Karachi in May 2011, have renewed international anxiety about the safety and security of Pakistan’s nuclear arsenal.
This article addresses several worrying trends in Pakistan that are coming together to suggest that the safety and security of nuclear weapons materials in Pakistan may very well be compromised at some point in the future.
The Growing Challenge of Securing Pakistan’s Nuclear Arsenal
In recent months, a variety of media sources have reported a significant escalation of nuclear weapons production by Pakistan. According to some of these sources, Pakistan has been building between 12 and 15 nuclear weapons a year, effectively doubling the size of its nuclear arsenal during the past three to four years to around 100 nuclear weapons. More disconcerting, Pakistan is engaged in a rapid expansion of its fissile material production through two new reactors, the Khushab II, thought to be operating in some form since 2009, and Khushab III, which has been under construction since 2005-2006 and is likely to come on-stream around 2013-2014. There is further evidence from the respected Washington-based Institute for Science and International Security that a fourth Khushab reactor may also be under early phase construction. Intended primarily to offset rival India’s conventional military advantage, the open-ended escalation of Pakistan’s nuclear weapons production explains why Pakistan has led the opposition to the international Fissile Material Cut-Off Treaty (FMCT), a treaty which would cap fissile material stockpiles.
Aside from the intricate politics of international arms control, the steady rise in the size of Pakistan’s nuclear arsenal presents the rather more prosaic, though arguably more serious, challenge of ensuring the physical security of an ever increasing number of nuclear assets. This is not a simple matter. Safeguarding 100 weapons is a significantly greater challenge than safeguarding 50 weapons because strategic and operational realities require that those weapons are dispersed and that dispersal locations are adapted to the complex requirements of safely and securely storing nuclear weapons in various degrees of operational readiness. As Pakistan’s nuclear arsenal grows in the years ahead, these challenges will multiply.
As many as 70,000 people in Pakistan reportedly have access to, or knowledge of, some element of the Pakistani nuclear weapons production, storage, maintenance, and deployment cycle, from those involved in the manufacture of fissile material, through those engaging in nuclear weapons design, assembly and maintenance, to those who transport and safeguard the weapons in storage and would deploy the weapons in crises. That number will also rise steadily as the size of the nuclear arsenal grows.
This figure is important because of the complex and highly polarized debates about nuclear weapons safety and security in Pakistan. All sides of that debate agree that Pakistan has, with considerable U.S. assistance, put in place a range of robust measures to seek to assure the safety and security of its nuclear weapons. The consensus breaks down, however, on the issue of whether these measures provide adequate safety and security for Pakistan’s nuclear arsenal. The Pakistan Army, which has overall control of the weapons, and Pakistan’s government argue forcefully that they do, although even they have recently moderated their statements of reassurance. Critics point to a number of vulnerabilities that place these reassurances in some doubt. These vulnerabilities boil down to three core concerns: a) that the physical security of nuclear weapons—across the weapons cycle—may not be robust enough to withstand determined terrorist assault; b) that among the estimated 70,000 people with access to the nuclear weapons cycle, some may be willing to collude in various ways with terrorists; c) that the threat extends beyond terrorists gaining access to complete and viable nuclear weapons, and include the immense political and security implications of terrorists gaining access to fissile material, nuclear weapons components, or penetrating nuclear weapons facilities.
A July 2009 article in the CTC Sentinel explained in detail the robust measures Pakistan has established to assure the safety and security of its nuclear weapons. It argued that terrorists have shown themselves able to carry out violent attacks at facilities that were reliably identified as having a nuclear weapons role. These facilities include the military complex at Wah, suspected to be involved in the manufacture of nuclear weapons parts; Kamra, suspected to be the designated base for the dispersal of nuclear assets in a crisis; and Sargodha, suspected to be a storage facility for nuclear delivery systems.
In none of these cases, however, were the terrorist attacks themselves aimed at penetrating the bases or at seizing nuclear assets; rather, they were mass casualty bomb attacks that took advantage of the fact that Pakistani security personnel were concentrated and relatively static at base entry points as they waited to go through security barriers. Some analysts criticized the article, arguing that: a) terrorists in Pakistan had never shown themselves capable of penetrating high security bases; b) that the secrecy of Pakistan’s nuclear weapons sites would ensure that terrorists could not know the locations of the weapons. Within a few months, the validity of both these counterarguments would be seriously undermined when Pakistani militants penetrated the Pakistan Army’s General Headquarters (GHQ) in Rawalpindi.
The Attack on Pakistan’s Army Headquarters
On October 10, 2009, Tehrik-i-Taliban Pakistan (TTP) and Lashkar-i-Jhangvi militants staged an audacious attack on the Pakistan Army’s GHQ in Rawalpindi, arguably one of the most secure military complexes in Pakistan, housing within its sprawling campus not only the chief of army staff, but also many of Pakistan’s most senior military commanders, including the director-general of the Strategic Plans Division (SPD) and the director-general of Strategic Forces Command (SFC)—Pakistan’s two most senior operational nuclear commanders.
The modalities of this attack add up to a virtual blueprint for a successful attack on a nuclear weapons facility:
– the penetration of layers of security checkpoints, barriers, and obstacles on the approach to the sensitive military site;
– the terrorist use of army uniforms and—according to some reports—a military vehicle with appropriate license plates, and forged ID cards, to deceive checkpoint personnel;
– the use of a safe house relatively close to the target site for several weeks before the operation to allow the build-up of a detailed intelligence picture;
– the use of a “sensitive” map (or maps) of the GHQ to allow detailed operational planning. The use of this map (or maps) point to one of two main possibilities: either that the attack had inside help, or that this kind of sensitive information is poorly controlled by the Pakistan Army/ISI;
– use of the kind of weaponry—small-arms, grenades and suicide vests—which allow final tier barrier defenses to be penetrated;
– use of tactics that allow final tier barriers to be penetrated: grenades and/or suicide detonations at entry points which then allow penetration by follow-up commando-style groups;
– use of diversionary tactics: attacking one gate first to draw off and weaken the defenses at a secondary entry point, perhaps closer to the main objective.
In all, at least 10 terrorists were involved in the operation, with four attacking the first gate, and a further six attacking the second gate. The terrorists gained entry to the complex where they took at least 40 people hostage. It took the Pakistan Army’s elite commandos, the Special Service Group (SSG), more than 20 hours to kill or capture all of the militants and free most of the hostages. Two civilians, seven Pakistani soldiers and five SSG commandos were killed in the raid.
In the months that followed the assault, several other disturbing aspects about the attack emerged. Among these was the assertion that intelligence about the attacks had been known to Pakistan’s Punjab government well before October 10 and that this intelligence had even been published in two Pakistani newspapers, The News International and The Daily Jhang, four days before the attack, but had been ignored by the Pakistan Army and ISI. It also emerged that the terrorists had, ironically, almost certainly learned their tactics from the SSG, which had trained earlier generations of Pakistani/Kashmiri militants in similar tactics for operations against India. In addition, there was a concerted effort by the Pakistan Army and ISI to manipulate the media reporting of the attacks, forcing several private TV channels temporarily off the air, contradicting or retracting certain details, and seeking to play down the significance of the assault.
The second set of features of the attack relate to secrecy, and they weaken the argument that Pakistan can ultimately rely on concealment to protect its nuclear assets. The use of “sensitive” maps in the attack, the time and proximity to conduct intelligence gathering, the level of knowledge of details such as uniforms, military plates, and possibly ID cards, point to a high level of terrorist knowledge of sensitive military information and protocols, whether through insider help or not. Furthermore, detailed knowledge of Pakistan’s security force movements and modus operandi has been a consistent feature of terrorist actions in Pakistan for many years, from the repeated assassination attempts against former President Pervez Musharraf, at least one of which included the insider involvement of Pakistani military officers, through the targeting of the ISI headquarters and vehicles, to the murders of senior military figures.
Pakistan’s Nuclear Security at Risk
In this context, given that nuclear weapons and delivery systems demand construction and other visible physical necessities (such as road widening, unusual levels of security, and bunker construction), and given that the growth of the Pakistani nuclear arsenal will significantly expand the construction of nuclear weapons infrastructure and the number of individuals with nuclear-related roles, it is simply not possible that the location of all of Pakistan’s nuclear weapons can remain unknown to terrorists in perpetuity.
As evidence of this, on August 28, 2009, the U.S. Federation of American Scientists published the first open source satellite imagery of a suspected Pakistani nuclear weapons storage facility near Masroor airbase outside Karachi. Within its perimeter walls, the satellite image shows three potential storage bunkers linked by looping roads. The fact that this image is available online, and that the unusual configuration of the base is clear, argues strongly that knowledge of the location of at least some nuclear weapons storage and other related facilities has reached terrorists in Pakistan.
As the number of nuclear weapons facilities grows, and the number of those with access to nuclear weapons or related components rises, the complex challenge of assuring the security of nuclear weapons and nuclear weapons components will become ever more difficult. Terrorist groups have now shown themselves capable of penetrating even the most securely defended of Pakistan’s military bases and of holding space within those bases for many hours even against the elite SSG, more than enough time with the right equipment and sufficient numbers to carry out terrorist acts with enormous political or destructive pay-off, from video broadcasts with the attention of the world’s media, through potentially destroying by explosions nuclear weapons or materials and the creation of a radiological hazard, to the possibility of the theft of nuclear weapons components or materials for subsequent terrorist use.
Indeed, on May 22-23, 2011, only about 15 miles from the suspected nuclear weapons storage facility near Masroor, a major terrorist attack targeted the naval aviation base at PNS Mehran in Karachi. Early reports suggest that between six and ten terrorists stormed the high security base from several entry points, that they had knowledge of the location of intruder detection cameras that they were able to bypass, and that they penetrated deep inside the base before using rocket-propelled grenades, explosives and small-arms to destroy several aircraft and take hostages. It took the base security and additional Pakistan Army rangers and commandos more than 18 hours to end the siege. At least 13 people were killed.
A frontal assault of this kind on nuclear weapons storage facilities, which are the most robustly defended elements of Pakistan’s nuclear weapons cycle, is no longer an implausible event. The successful location and penetration of such a site by terrorists, even if they were ultimately unsuccessful in accessing nuclear assets, would itself be a transformative event both in terms of the U.S.-Pakistani nuclear relationship and in terms of international anxiety about the security of Pakistan’s nuclear weapons. Such an assault would also critically undermine Pakistan’s reassurances about the security of nuclear weapons elsewhere in the weapons cycle, particularly in transit. As the number of Pakistani nuclear weapons inexorably continues to rise, and as the nuclear weapons security challenges thereby steadily multiply, the odds that Pakistan’s nuclear weapons security will eventually be compromised continue to rise.
Professor Shaun Gregory is Director of the Pakistan Security Research Unit at the University of Bradford, UK. He has published widely on nuclear and security issues in Pakistan and advises many governments, their agencies, and international organizations.
 Karen DeYoung, “New Estimates Put Pakistan’s Nuclear Arsenal at More than 100,” Washington Post, January 31, 2011.
 “Pakistan Appears to Expand Nuclear Site – Report,” Reuters, February 10, 2011. For fuller details of this expansion, see the Institute for Science and International Security’s website at www.isis-online.org.
 If concentrated in too small a sub-set of locations, the weapons constitute a lucrative and vulnerable set of targets.
 This figure includes the 8,000-10,000 staff of Pakistan’s Strategic Plans Division and 7,000-8,000 scientists of whom 2,000 are reported to have “critical knowledge” of Pakistan’s nuclear weapons. The latter figure was originally attributed to the director of the Strategic Plans Division, Lieutenant General Khalid Kidwai, in January 2009. For details, see David E. Sanger, “Obama’s Worst Pakistan Nightmare,” New York Times, January 8, 2009. It also includes up to 18,000 troops reported to guard the nuclear assets. For details, see Andrew Bast, “Pakistan’s Nuclear Surge,” Newsweek, May 15, 2011. For a development of these issues, see Christopher Clary, “Thinking About Pakistan’s Nuclear Security in Peacetime, Crisis and War,” Institute for Defence Studies & Analysis, September 2010; Matthew Bunn, “Securing the Bomb 2010,” Belfer Center for Science and International Affairs, Harvard University, April 2010.
 It is interesting that the official Pakistani narrative has moved from a comprehensive insistence that Pakistan’s nuclear weapons are totally safe and secure, to a rather less absolute (and more realistic) acceptance that Pakistan’s nuclear weapons are “at least as safe and secure as those of any other nuclear country.” This is an important change because all nuclear weapons states that have put information in the public domain have accepted that accidents, breaches of security, and unintended events are an inevitable part of operating a nuclear arsenal. None of these states, however, face the terrorist threat level that confronts Pakistan. For senior Pakistani articulation of the latter position, see, for example, “Strategic Assets Are Safe, Says FO,” Dawn, November 13, 2007.
 There must also be a risk that among this number are terrorists or their sympathizers who have applied for jobs to gain access to part of the weapons cycle, with the smuggling of fissile material a key vector of concern. For an insightful analysis of these issues, see Brian Cloughley, “Fission Fears,” Jane’s Intelligence Review, April 2011.
 It must be conceded here that the claimed roles of the sites at Wah and Kamra are robustly contested by some in Pakistan and cannot be definitively resolved on the basis of unclassified sources. Wah is Pakistan’s main conventional weapons production facility and home to at least 14 separate complexes dealing with technologies including explosives, heavy artillery ammunition, steel and alloy, propellants, and weapons manufacture. These are precisely the subset of technologies necessary to manufacture and assemble the non-nuclear components of nuclear weapons (warhead cases, conventional explosive triggers, etc.) into which the fissile core can be fitted. In the absence of comparable alternative facilities in Pakistan, Wah remains the most likely location for the manufacture and assembly of nuclear weapons parts. What is less clear is whether fissile material is enriched or otherwise worked at Wah or whether fissile material is brought to Wah for final assembly. For details on these matters, see www.globalsecurity.org/wmd/world/pakistan/wah.htm and www.cns.miis.edu/reports/pdfs/pakistan.pdf. Pakistan’s Air Weapons Complex Kamra, close to Wah, is reported to have a role in air-delivered nuclear weapons and to be a dispersal site for aircraft armed with nuclear weapons in crises. For details, see www.globalsecurity.org/wmd/world/pakistan/kamra.htm and Paul Kerr and Mary Beth Nikitin, “Pakistan’s Nuclear Weapons: Proliferation and Security Issues,” Congressional Research Service, January 13, 2011, available at www.fas.org/sgp/crs/nuke/RL34248.pdf. There was no similar contestation of Sargodha’s possible role as a storage facility for nuclear ballistic missiles. For details on Sargodha, see www.globalsecurity.org/wmd/world/pakistan/sargodha.htm.
 The sources used in the article make entirely clear the nature of the attacks to which the argument was referring. See Shaun Gregory, “The Terrorist Threat to Pakistan’s Nuclear Weapons,” CTC Sentinel 2:7 (2009).
 This list has been put together from interviews and a range of sources, including: “Press Review: Rawalpindi Attack,” BBC, October 12, 2009; Shahid Rao, “Terror Attack on GHQ,” The Nation, October 11, 2009; Pakistan’s Inter Services Public Relations, press releases, May 2011; Hassan Abbas, “Deciphering the Attack on Pakistan’s Army Headquarters,” Foreign Policy, October 11, 2009.
 “GHQ Attack Report Published in Daily Jhang, The News on 5th October,” The News International, October 11, 2009.
 These are some of the reasons it is difficult to be definitive about all of the attack details.
 “Air Force Officers Held for Attack on General Musharraf,” Daily Telegraph, November 5, 2006.
 “Bombers Hit Pakistan Spy Agency,” BBC, November 13, 2009.
 Those killed include Lieutenant General Mushtaq Baig, at the time Pakistan’s surgeon general, blown up at a road junction in February 2008, and Major General Amir Faisal Alvi, former head of the SSG, gunned down on his way home. See respectively “Rawalpindi Suicide Blast: Kills 8 with Pakistan Army Surgeon General,” Pakistan Times, February 25, 2008; Syed Shoaib Hasan, “Top Pakistan Ex-Commando Killed,” BBC, November 19, 2008.
 Nor should it be doubted that some terrorists at least continue to seek nuclear weapons or components. For an interesting discussion, see Abdul Bakier, “Jihadis Discuss Plans to Seize Nuclear Assets,” Terrorism Monitor 7:4 (2009).
 To view the U.S. Federation of American Scientists’ document, see www.fas.org/blog/ssp/category/pakistan.