IN THE HUNT FOR MEMBERS of al-Qaeda and other terror groups, Western intelligence agencies use special software that sifts through enormous quantities of diverse data to map “non-obvious relationships” among people, places, behaviour and ideas. If, for example, two Saudi men have phoned a radical Afghan cleric and twice stayed in the same Islamabad hotel when the cleric was also in the city, the three may be working together. This effort to sort out “who’s who in the zoo”, in the words of Bob Griffin, head of i2, a British developer of the software, identified a certain man in Pakistan whose atypical behaviour and social ties were particularly intriguing.
The man repeated certain behaviours in the days or hours before the release of Osama bin Laden’s communiqués – perhaps, for example, travelling to meet a non-relative from an apparently different walk of life and with whom he neither communicated electronically nor interacted publicly. By looking for possible linkages in data from diverse intelligence sources, “network analysis” software, as it is known, had apparently helped identify one of bin Laden’s couriers, says Mr Griffin, whose firm’s software was used in the manhunt.
Part 4 explores technology’s impact on intelligence and spycraft. The technologies are diverse. Spy drones will shrink to about the size of a large insect. At some American military checkpoints video technology analyses body movements to detect hostile intent. For cargo-scanning, a sort of X-ray vision can be obtained using a hydrogen isotope to bombard a heavier form of the gas. Yet the burgeoning field of network analysis best illustrates how technology is driving transformations in the intelligence world as a whole. To begin with, many of the technologies examined in this section, no matter what their stripe, will increasingly gather or process information to feed network-analysis systems.
Breakthroughs in radar instrumentation, for example, are making it easier to discover tunnels and bunkers. And American military software that analyses aerial video of traffic can detect suspicious vehicles (such as those that pause near a building that is later attacked or return to their starting point with no one having left or entered the vehicle). The information from both radar and video is useful. But it becomes far more valuable when woven together with network analysis. This might reveal that, say, a truck flagged as suspicious is parked close to a newly discovered smuggling tunnel near a border and that the driver appears linked (through phone calls or other means) to a member of an armed group.
Advances in network analysis underscore improvements in spycraft technologies as a whole. During the 2003 hunt for Saddam Hussein, it was difficult to integrate network-analysis data from different computer systems. So American army analysts partly resorted to using string to link photographs and notes tacked to the wall of an operations room in Iraq, says Ian McCulloh, an instructor at the West Point military academy who conducts network-analysis workshops for intelligence officials. Stephen Borgatti, a University of Kentucky expert who has helped the Department of Defence develop network-analysis software, says that not long ago intelligence analysts commonly covered walls with sticky notes and pictures and “just stared at it until something came to them”.
Drones both produce and require copious intelligence. Accordingly, investment in spy technologies, including network analysis, has boomed along with the use of drones. In the five years before 2009, American drone strikes killed at least 215 militants in Pakistan. The death toll leapt to at least 1,012 militants in the following two and a half years, according to the conservative estimates in one tally. Network analysis helped identify many of those targets. The tool will become even more formidable as other spy technologies leave the lab. For example, devices that analyse the wavelengths of infrared light reflected by airborne particles may soon allow drones to sniff out explosives or clandestine bomb-making facilities.
Intelligence technologies often fuel privacy worries, perhaps none more so than network analysis and its cousin data mining. Massive amounts of data are collected about people’s whereabouts, purchases and activities. Is the loss of privacy worth it?
Many would say yes, at least to combat serious threats. Worldwide, more than 25 agencies use network analysis to detect trafficking in nuclear materials and expertise, according to one estimate. The technology works well, notes a Frenchman who leads a firm that calibrated a model using data on the nuclear-smuggling network of Abdul Qadeer Khan, a Pakistani. Thanks to university and research-lab records, he says, most nuclear experts are known and accomplices generally share strong ties (needed to develop trust) that the software makes apparent.
By and large, the public has also accepted efforts to flush out lesser wrongdoers by analysing all manner of legal but potentially suspect behaviour (including web browsing). To detect fraud, software developed by an American firm, SAS, can link Internal Revenue Service tax records to credit-card data, property deeds and much more. The police department of Richmond, Virginia, uses network analysis with data from Facebook, MySpace and Twitter to better dispatch police officers to the parties most likely to become rowdy. A computer scientist at Britain’s Cambridge University likens the general acceptance of reduced privacy to a “boiled frog” that did not jump out of a saucepan because it was heated gradually.
Upcoming intelligence technologies will go well beyond Western efforts to “make every soldier a sensor”, says Jared Freeman, the top technologist at Aptima, a Massachusetts company. With US Department of Defence funding, Aptima is developing a handheld network-analysis system called NETSTORM, which will rapidly convert raw information from the field into processed tactical intelligence. A soldier or police officer might, for example, upload the name, occupation and phone number of a young villager being questioned. Quickly returned intelligence might reveal that, say, the villager had received a phone call from a wealthy and unrelated foreigner who had exchanged e-mails with a teenager who became a suicide bomber.
In matters of war and peace, intelligence has been enormously valuable for millennia. But the electronic age is making intelligence far easier to gather, process and use. Electronics have also made it easier to defeat some sophisticated weaponry with intelligence alone.
In the run-up to the first Gulf war, America obtained intelligence on the technical specifications of Iraqi air-defence systems imported from the Soviet Union, says David Kay, formerly the top UN weapons inspector in Iraq. Based on the information, adjustments were made to coalition electronic-warfare systems that rendered Iraq’s anti-aircraft missiles “extremely ineffective”, says Siemon Wezeman, an arms-trade expert at the Stockholm International Peace Research Institute, a Swedish think-tank.
The trick is known as “delinking” a missile. Efforts to obtain this sort of intelligence in the electronic age have spawned a booming underground trade, says an official at the Swedish Defence Research Agency, a defence-ministry body in Stockholm. The agency, he says, continuously gathers and shares this sort of technical information with allies. Rarely is it sold; instead, countries swap such secrets or pass them on to buy goodwill, the official says.
The use or development of spy and intelligence technologies is unlikely to grab as many headlines as the use or sale of weaponry. But, as the following articles show, the myriad technologies of modern spycraft are nonetheless dramatically reshaping efforts to keep the peace or prevail in conflict.