Pakistan's nukes are safe. Maybe.

By Vipin Narang

An excellent series of recent articles on the subject by Shaun Gregory, Rolf Mowatt-Larssen (a former director of intelligence at the Department of Energy), and Brig. Gen. Feroz Hassan Khan (Ret.) assess the very grim threat of Pakistan losing control over its 60-warheads-and-growing nuclear weapons arsenal. Given the lack of publicly available data on this critical issue, such articles by extremely knowledgeable scholars and practitioners represent some of the best information we have on realistic threats to Pakistan's nuclear arsenal.

Gregory's article has gotten some recent attention for noting that there have worryingly been several attacks at the perimeter of bases that may house nuclear components, though U.S. intelligence officials are quick to point out that there is little reason to believe that nuclear assets were ever at risk. So what are the primary risks to the security of Pakistan's nuclear arsenal?

In answering this question, it is important to differentiate between the various organizations involved with Pakistan's nuclear weapons, and where and when nuclear assets are more or less vulnerable to internal and external threats. The bigger threat is probably not the Army losing control of nuclear assets, but rather insider-outsider collusion or diversion of nuclear material from the civilian nuclear agencies during either the production phase or transfer to Army locations.

The good news is that once the Pakistani Army takes custody of nuclear assets, the threat of terrorists successfully boosting a warhead or fissile cores -- either through direct attack or facilitated by insiders -- is reassuringly low. The Pakistani Army has every incentive to ensure firm control over the country's nuclear assets since, should they be lost or stolen, there would literally be hell to pay.

A combination of security measures instituted by the Strategic Plans Division, headed by Lt. Gen Khalid Kidwai (Ret.), has increased confidence that the Pakistani Army has firm control over nuclear assets, particularly during peacetime with India, when Pakistan's nuclear weapons are believed to be kept in component form with warheads disassembled and separated from the missiles that would deliver them, in highly secret fixed locations that are easier  to protect through concentric rings of security forces (believed to include the ISI, regular Army, and dedicated elite forces to protect nuclear assets).

Even in the unlikely event that Bruce Riedel's nightmare scenario of a jihadist takeover of the Pakistani government unfolds, there is little reason to believe that the Army would give up or lose control of nuclear weapons -- indeed, there is little evidence that the Army has relinquished control even to Pakistan's current civilian government. The highly professional nature of Army units charged with guarding Pakistan's nuclear assets, procedural protections that require at least the "two-man rule," the SPD Personnel Reliability Programme (which monitors the loyalty and mental states of military personnel), and the ability to protect fixed locations against most realistic terrorist threats suggest that in peacetime the Pakistani nuclear arsenal that is in the custody of the Army ought to be relatively secure.

Now, for the potentially bad news. As Gregory, Khan, and Mowatt-Larssen all suggest, the primary risk to the Pakistani Army's ability to safely secure nuclear assets in its custody would likely be during crisis scenarios -- either against India or due to a perceived Western threat to the integrity of Pakistan's arsenal -- that might cause Pakistan to move to a higher state of nuclear readiness. If the Army feels compelled to rapidly disperse or relocate nuclear components and loses the defensive advantage of protecting them in secure fixed locations, insider foreknowledge of movements and the loss of centralized control could increase the probability of theft or loss.

This threat could be magnified if the Pakistani Army assembles warheads before moving them for procedural or technical reasons, thereby removing some -- if not all -- safeguards preventing unauthorized or accidental detonation. There is some ambiguity about the so-called Pak-PALs (Permissive Actions Links, which are basically locks for nuclear warheads) that Kidwai claims Pakistan has developed to prevent unauthorized or accidental use. Modern U.S. PALs, for example, are digitally integrated into the firing system of a fully assembled weapon, so developing PALs for disassembled weapons may perhaps involve only rudimentary physical locks to prevent assembly of the weapon without a proper code (either colocated with the assets or held centrally) that are easier to bypass.

The removal of already-weak safeguards as Pakistan moves to a higher state of nuclear readiness would be consistent with its nuclear posture of credibly threatening the early first use of nuclear weapons against Indian forces to deter an Indian conventional attack, but it increases safety and security risks.

There are some indications, though no clear publicly available evidence, that Pakistan has moved or readied nuclear assets several times, and only in response to external threats: in the 48 hours after 9/11 when General Musharraf feared that the U.S. might attack Pakistan and perhaps during 2002 when Pakistan feared that a major Indian conventional attack was imminent during the Parakram crisis. So although crisis scenarios may pose the greatest risk to the Pakistan Army's ability to securely guard Pakistan's nuclear weapons, the empirical record suggests that Pakistan has only alerted forces in supreme emergencies when the state's survival was potentially threatened by external threats, so this threat -- while very real -- should not be overstated.

Pakistan's nuclear arsenal is also vulnerable at several points before the Army takes control of assets from the civilian nuclear infrastructure in fixed locations. Thousands of civilian personnel are responsible for the production of Pakistan's nuclear weapons components. In recent years Pakistan has established a Human Reliability Programme (HRP) to regularly monitor civilian nuclear personnel, ensuring that there are no extremists within the system, and has also established a Nuclear Security Action Plan to establish best practices and deal with recovering orphaned radioactive sources (mostly contaminated metals).

There are a few scenarios that could pose risks to Pakistan's nuclear weapons, aside from the obvious risks of direct attacks on civilian nuclear installations or radical insiders within the organization who might transfer nuclear weapons to terrorist organizations (e.g. a repeat of the UTN episode or an A.Q. Khan-like figure willing to do business with al Qaeda or the Taliban); security around nuclear installations is robust and tight, though nothing is ever totally impenetrable, while the PRP/HRP is tasked with ensuring that the latter scenario does not unfold.  

The first risk is the threat of a diversion of fissile material at the production stage by one or more Pakistani nuclear scientists, either at once or slowly over time, with the latter being more difficult to detect. Given the accounting uncertainties in even the world's most secure nuclear facilities under IAEA safeguards, it might be impossible to detect the diversion of a few kilograms of fissile material accumulated over long periods of time, particularly as Pakistan is ramping up production of the weapons-grade material (both plutonium and HEU) needed to make nuclear weapons.

While protection and accounting in Pakistan's nuclear establishments have reportedly improved since 2001, perhaps partly with U.S. assistance (though much of this remains highly classified and it is not clear how much Pakistan trusts or would accept U.S. assistance), it was considered rudimentary prior to that and it is therefore a daunting task to ensure that 100 percent of Pakistan's fissile material produced over the past 20 years is accounted for, since accounting uncertainties of just a few tenths of a percent could be sufficient to develop a radiological or nuclear device -- and no regulatory authorities may even realize the material is missing.

The second point of vulnerability is during the transfer of nuclear assets from civilian organizations to Army locations. Mowatt-Larssen writes that Pakistan "transport[s] and deploy[s] weapons clandestinely rather than in convoys that have a stronger, highly visible security profile" which could "paradoxically ... backfire in the event a malicious insider gained access to locations of weapons storage sites, transportation routes, and similar insider information, especially because more moving parts are involved in assembling weapons when they are being deployed. In such a case, there would be fewer barriers between an outside group and the bomb."

That is, Pakistan relies on less-guarded, secret transfers instead of heavily armed convoys, which means that an attack that either knowingly (through insider collaboration) or by chance targeted a nuclear transport would have a higher probability of successfully boosting nuclear material. This could be a particularly attractive mode of attack since terrorist organizations might be able to successfully steal nuclear assets with only the assistance of a lone malicious insider who had foreknowledge of transport times and routes.

The key point here is that there are different organizations responsible for the production and stewardship of Pakistan's nuclear weapons and each one presents its own points and scenarios of vulnerability. The Pakistani Army may face acute challenges ensuring centralized command and control of nuclear assets during times of crisis. And with respect to the civilian nuclear agencies, highly accurate accounting and controls, rigorous screening procedures, and secure transfer of materials to and by the Army are crucial to ensuring that Pakistan's nuclear assets are not vulnerable to corrupt insiders or external terrorist threats.

Vipin Narang is a Ph.D. candidate in the Harvard University Department of Government and a research fellow at Harvard's Belfer Center for Science and International Affairs.

AFP/Getty Images