Category: Security

Security is immunity from the will of others.

Palantir and data analysis

Writing in The Guardian, Jacques Peretti has compiled an interesting summary of the technological capabilities and government-to-business relationships of Palantir, a secretive technology company focused on identifying patterns within large data sets and making them accessible to people without specialized training.

With sensors getting cheaper all the time, the tricky part of ubiquitous surveillance isn’t collecting the data. It’s making it intelligible and applicable. These kinds of powerful data linkage and analysis tools also undermine common-sense expectations and procedures for the protection of privacy. No human being might be able to look at a set of large supposedly-anonymized databases and pick out individuals, but it’s increasingly within the scope of what can be routinely done with computers.

Grating coupler arrays as cameras

A recent Economist article describes a novel camera design with the promise to be far thinner than those that exist now, with some novel features:

Not only do Dr Hajimiri’s cameras have no moving parts, they also lack lenses and mirrors—in other words, they have no conventional optics. That does away with the focal depth required by today’s cameras, enabling the new devices to be flat.

…

To mimic the image-making role of the optics in conventional cameras, the OPA manipulates incoming light using electrons. Dr Hajimiri compares the technique to peering through a straw while moving the far end swiftly across what is in front of you and recording how much light is in each strawful. In the OPA this scanning effect is created by manipulating the light collected by the grating couplers electronically, using devices called photodiodes. These place varying densities of electrons into the amplified light’s path through the OPA, either slowing it down or speeding it up as it travels. That shifts the arrival times of the peaks and troughs of the lightwaves. This “phase shifting” results in constructive interference between waves arriving from the desired direction, which amplifies them. Light coming from other directions, by contrast, is cancelled through destructive interference. Change the pattern of electrons and you change the part of the image field the OPA is looking at. Scanning the entire field in this way takes about ten nanoseconds (billionths of a second).

…

To zoom in for a close-up, the device selects a specific part of the image and scans it more thoroughly. To zoom out for a fish-eye, it scans the entire optical field, including light from the edges of that field. To change from zoom to fish-eye takes nanoseconds.

Doubtless, such cameras will have some interesting applications. Unfortunately, that will certainly include further entrenching the surveillance state — increasingly using devices too small to see.

Security of Pakistani nuclear weapons

Nevertheless, the exact nature of launch authorization procedures is ambiguous. Several sources refer to a system of two separate codes—one civilian and the other military—amounting to a “dual key” system. However, several authoritative accounts mention a three-man rule. In particular, the code to arm a weapon can only be inserted in the presence of three persons. It is possible that a two-man rule is adopted for movement of warheads and a three-man rule is adopted for employment authorization. According to Pakistani planners, the number of persons involved varies “for technical reasons”—three at some points in the chain of command, two at other points.

Pakistan is not explicit about its arrangements for weapons security, but it has developed physical safety mechanisms and firewalls both in the weapon systems themselves, as well as in the chain of command. No single individual can operate a weapon system, nor can one individual issue the command for nuclear weapons use. The NCA command and control system ensures that weapons can be operationally ready on short notice, yet unauthorized arming and/or use never takes place.

Pakistan does not keep its nuclear weapons on hair-trigger alert. The nuclear weapons are few in number and probably kept in disassembled form; their components are reportedly stored separately, at dispersed sites. Keeping the weapons in a disassembled form, along with the use of authorization codes, reduces the risk of capture or unauthorized use. Naturally, there is considerable uncertainty about the location of Pakistani nuclear weapons and about procedures for actual use. After September 11, Pakistan ordered a redeployment of the country’s nuclear arsenal to at least six secret new locations, according to one account. Fissile materials are obviously stored in secret locations; probably in initial stages they are near installations such as Kahuta or Khushab, or close to Rawalpindi. Additionally, from a security standpoint sensitive material sites are carefully chosen, in safe areas and within quick reach of designated rapid reaction forces, which are specially trained and operate under command of the security division of SPD. Although Pakistan’s system is not as sophisticated as the U.S. permissive action links (PALs), it is deemed reliable enough to preclude unauthorized arming or launching of its nuclear weapons.

Dummy locations are also reportedly employed to minimize the risks of destruction or capture. SPD Head Lieutenant-General Khalid Kidwai, in a lecture at the U.S. Naval Postgraduate School in October 2006, clarified that “no delegation of authority concerning nuclear weapons is planned.” The conclusion, therefore, is that centralized control is retained by the NCA at the Joint Services Headquarters. Beyond this clarification, operational control plans cannot be made public by any nuclear state and thus remain a national security secret, as was the case with the United States and other nuclear powers during the Cold War.

Khan, Feroz Hassan. Eating Grass: The Making of the Pakistani Bomb. Stanford University Press; Stanford. 2012. p. 331–2

Chinese aid to Pakistan’s nuclear weapons program

Western sources claim that China had provided Pakistan with fissile material in exchange for centrifuge technology assistance. Zia-ul-Haq hoped to exploit the close relationship with the Chinese further in order to protect Pakistan from potential preventative attacks… [T]he impact of Israeli attack on Osirak and the crash of the centrifuges in 1981 forced Zia-ul- Haq to realize that the nuclear program was vulnerable not just to preventive strikes but also to natural calamities. Zia-ul- Haq then dispatched Lieutenant-General Syed Zamin Naqvi and A.Q. Khan to request bomb-grade fissile materials and bomb designs. Their visit bore fruit as Pakistan then received the Chinese CHIC-4 weapon design along with 50 kilograms of HEU in 1981, material sufficient for two bombs. A.Q. Khan confirmed in a purported 2004 letter to his wife, “The Chinese gave us drawings of the nuclear weapon, gave us 50 kg of enriched uranium, gave us 10 tons of UF6 (natural) and 5 tons of UF6(3%).”

According to A.Q. Khan’s accounts, the Chinese nuclear material was kept in storage until 1985. When Pakistan acquired its own uranium enrichment capability and wanted to return the fissile material, China responded that “the HEU loaned earlier was now considered as a gift … in gratitude” for Pakistan’s help with Chinese centrifuges. It was then that KRL “promptly fabricated hemispheres for two weapons and added them to Pakistan’s arsenal.”

Khan, Feroz Hassan. Eating Grass: The Making of the Pakistani Bomb. Stanford University Press; Stanford. 2012. p. 188 (typographical inconsistencies in original)

Tactics of rebellion in Egypt

Accounts of the protests that brought down the Mubarak government stressed the role of new internet-based social media, which helped organisers and supporters plan the protests. The critical event in toppling the regime, however, was the initial seizure of Tahrir Square on 25 January – a development in which the social media functioned partly as a decoy. Knowing that the security forces would use violence to break up any attempt to occupy the square, the organizers used social media to plan protests at twenty sites in working-class districts of the city, hoping to strain the security forces by dispersing them to multiple locations, while drawing large crowds that would increase the chance of breaking through security cordons and linking up at Tahrir Square. They planned one additional gathering, in Bulaq al-Daqrur, a working-class neighbourhood close to the centre of the city, with an industrial workforce employed in a nearby cigarette factory and in railway yards. They avoided announcing this gathering over the internet, allowing a crowd of several hundred to gather without the presence of security forces. This was the group that marched to Tahrir, swelling to several thousand along the way, and seized the square, by which time the protest was too large for the armed police force to crush.*

* Footnote: Charles Levinson and Margaret Coker, ‘The Secret Rally that Sparked an Uprising‘, Wall Street Journal, 11 February 2011.

Mitchell, Timothy. Carbon Democracy: Political Power in the Age of Oil. Verso; London. 2013. p. 229

Automated voice impersonation

I’ve written before about some problems with biometric security: it seems convenient to be able to use facial recognition to log in to your computer, until you find your co-workers doing it with colour photocopies of your picture.

Computers aren’t the only context where we use biometrics for identification. “Don’t you recognize my voice?” has been used for decades for authentication over the phone, whether implicitly or explicitly. Now, we’re approaching the day when faking anybody’s voice and having it say anything you like is getting near.

Expect disruption on every level, from teens pranking each other to abusive harassers terrifying victims in new ways to more election-altering political fraud.

Quantum sensors and vulnerable submarines

A recent technology quarterly about new quantum innovations, published in The Economist, referred to a disturbing development in quantum sensor technology:

Military types are interested, too. “You can’t shield gravity,” says David Delpy, who leads the Defence Scientific Advisory Council in Britain’s defence ministry. Improved gravity sensors would be able to spot moving masses under water, such as submarines or torpedoes, which could wipe out the deterrent effect of French and British nuclear submarines.

So much of the present nuclear balance of power (such as it is) depends on ballistic missile submarines being essentially invulnerable by virtue of being impossible to locate. Reportedly, almost no crew members about an American boomer (as subs carrying nuclear missiles are known) know the precise location of the ship, and nobody on land has the information.

If states suddenly feel their subs are vulnerable, it risks two big effects. First, it raises tensions in a crisis. If states fear they will lose their seaborne second strike capability, they may be inclined to launch a nuclear attack earlier. Second, if the safest leg of the nuclear triad (along with bombers and land-based missiles) suddenly seems vulnerable, it’s likely they will assemble and deploy more weapons in more locations, wasting money and raising the risk of accidental or unauthorized use.

As with other emerging nuclear-related technologies like hypersonic weapons, it would be better for everyone if we could agree to prohibit sensors that threaten subs. Alas, states are rarely so cooperative or trusting.

Dictatorships and the coercive dilemma

Coercive institutions are a dictator’s final defense in pursuit of political survival, but also his chief obstacle to achieving that goal. This book argues that autocrats face a coercive dilemma: whether to organize their internal security apparatus to protect against a coup, or to deal with the threat of popular unrest. Because coup-proofing calls for fragmented and socially exclusive organizations, while protecting against popular unrest demands unitary and inclusive ones, autocrats cannot simultaneously maximize their defenses against both threats. When dictators assume power, then, they must (and often do) choose which threat to prioritize. That choice, in turn, has profound consequences for the citizens who live under their rule. A fragmented, exclusive coercive apparatus gives its agents social and material incentives to escalate rather than dampen violence, and also hampers agents from collecting the intelligence necessary to engaged in targeted, discriminate, and pre-emptive repression. A unitary and inclusive apparatus configured to address significant mass unrest, however, has much better intelligence capability vis-a-vis its own citizens, and creates incentives for agents to minimize the use of violence and to rely instead on alternative means of repression, including surveillance and targeted pre-emption. Given its stronger intelligence capability, the mass-oriented coercive apparatus is also better at detecting and responding to changes in the nature of threats than its coup-proofed counterpart, leading to predictable patterns of institutional change that are neither entirely path dependent nor entirely in keeping with the optimization predicted by rational design.

Greitens, Sheena Chestnut. Dictators and Their Secret Police: Coercive Institutions and State Violence. Cambridge University Press. 2016. p. 4-5

Sheena was in the Oxford M.Phil in International Relations program during the same two years as I was, and we both served on the executive of the Oxford University Strategic Studies Group.

Related: