criminal warrants (or Title III warrants), even when it comes to the interception of US citizens’ communications. More seriously, though, the FISC has authorised the establishment of mass surveillance programmes in terms of the amendments to FISA by the Amendment Act, and, according to the Snowden documents, the court had used a controversial section of the Patriot Act to approve the collection of the phone metadata records of Americans in bulk. In authorising the procedures for mass surveillance programmes, FISC had no concrete knowledge of their actual functionings, and as a result deferred to the intelligence agency applicants rather than second-guess them. Under these circumstances, it is hardly surprising that the FISC has been a poor check on official surveillance abuses.11
WHY WORRY ABOUT SURVEILLANCE?
Why should we be worried about surveillance? After all, governments argue, if people have nothing to hide, then they have nothing to fear: a saying whose origin is unclear, but which has been attributed to Orwell and, before him, the Nazi Minister of Propaganda, Joseph Goebbels. The problem with this argument is that it assumes that government motives in undertaking surveillance are pure: that is, they will not misuse these powers to spy on their political opponents and others whom they consider to be politically inconvenient, like investigative journalists intent on holding governments to account. Governments the world over have justified surveillance to their citizens as a necessary evil to counter criminal activities and threats to national security, especially those posed by terrorists. Yet, even in the wake of the terrorist attacks on the US on 11 September 2001, the ‘war against terror’ never remained a pure war against those who were responsible for their acts and their supporters. Some countries seized on the heightened global security measures to place political activists under surveillance. In fact, the expansion of security powers was under way even before 11 September. In New Zealand, for instance, the country’s definition of national security was extended to include the country’s economic and international well-being as far back as 1996.12 The UK government spy agencies MI5 (the domestic intelligence agency) and MI6 (the foreign intelligence agency) have spied on social movements calling for nuclear disarmament and trade unions, clearly considering them to be an ‘enemy within’, in spite of scant evidence of illegality.13 There are many well-documented cases of states, under the political direction of government officials, using their surveillance capacities to spy on activists who posed no real threat to national security; these practices have cast a pall on the supposed political neutrality of state surveillance practices and the uses to which they are put.14 These cases point to a broader truth about surveillance – namely, that it is not being used simply to fight crime and terrorism but to stabilise social relations more broadly by collecting information on those who may expose or challenge how society is currently organised. The purpose of surveillance in this broader context is to make people who are considered to be problem subjects legible to public and private power, for the purposes of determining their risk profiles, managing them as risks and even acting against the most extreme risks. This is particularly the case in relation to intelligence work, where broad definitions of what constitutes threats to national security are coupled with technologically advanced forms of surveillance, leading to state spies claiming the right to access more and more personal information.
There are many forms of intelligence-gathering, but the two that are of concern in this book are human intelligence and signals intelligence. Human intelligence (or HUMINT) is intelligence gathered by physical means, such as the infiltration of organisations, tailing of suspects or espionage, while signals intelligence (or SIGINT) is intelligence gathered from the surveillance of electronics networks, including telemetry intelligence, electronic intelligence and communications intelligence (or COMINT).15 Globally, intelligence agencies are being tempted to shift away from human intelligence and towards signals intelligence. Why is this so?
Technological advancements have reduced the cost of surveillance using electronic signals, making it easier for governments to place whole populations under surveillance. Previously, if governments wished to spy on their citizens, they would need to place them under physical surveillance – which could also be uncovered, possibly at great political cost to the government concerned – or need to use costly telephone-bugging equipment that could be detected if people knew what to look for. However, new information and communications technologies (ICTs) have allowed governments and private companies to gather vast amounts of information about people at a small fraction of the previous cost. They are able to collect and analyse huge databases detailing people’s travel habits, banking details, personal interests and reading habits, who their closest friends and associates are, and what they communicate about. For instance, cellphones can provide huge amounts of information about whom people communicate with, as well as location information about where they have travelled. Internet and communications companies are also busy storing information about people’s locations, search habits and communications with other people, to mine it for commercial uses. This information is called metadata, as it provides information about a person’s communications, but not the actual content. However, metadata can provide such rich information about a person’s habits, personal preferences and associations that it can be as revealing as, if not more revealing than, communications content. Yet there are few legal protections for metadata.
Other data-driven technologies are being used increasingly to monitor public spaces, for a range of national security, criminal justice and commercial reasons. CCTV cameras can monitor people’s movements, and increasingly these cameras are fitted with invasive technologies such as facial recognition, which allows a computer to identify a person from his or her facial features from a photograph stored in a facial database (known as ‘smart’ CCTV, which is something of a misnomer as once the data is transmitted over the internet to another location, the television system no longer operates on a closed circuit). Electronic tolling systems track drivers’ movements for the purpose of billing them for road usage. These schemes usually include automatic number-plate recognition systems, allowing tolling companies to recognise vehicle licence plates and match them to a database of vehicle owners. Biometric identifiers – or digital impressions of fingerprints, irises or voices – can also be used to identify people for surveillance purposes; this is a particularly invasive form of identification as it involves a person’s body. While many countries use biometrics for border control, some have attempted to use them to establish national identity systems. In countries like the US and the UK, these attempts have been met with mass opposition on the ground that they threaten privacy on a massive scale. More governments are also using unmanned assistance vehicles – colloquially known as ‘drones’ – to conduct surveillance of borders and monitor crime scenes and disaster areas. Drones can also be loaded with lethal payloads to kill opponents, and, controversially, lethal drone strikes became a distinctive feature of the Barack Obama administration.
These surveillance devices can gather vast datasets, or ‘big data’, which can be analysed using computer algorithms, or sets of rules or methods for the processing of data and for the purpose of establishing patterns. Search engines like Google mine people’s internet searches to analyse personal interests, and then on-sell this information to advertisers. But big data is being used increasingly by intelligence agencies as well. One of the challenges of living in the digital age is that we are not necessarily who our data suggests we are: we can develop ‘data doubles’, or digital identities that may bear little, if any, resemblance to who we really are and what we really think.16 These identities can then be used to profile people as terrorists or criminals, even if they are innocent. So, intelligence analysts may tell computers to analyse datasets for individuals who have searched for information on how to buy a gun, have spent time in Syria or Iraq recently – where ISIS established its strongholds – and have made enquiries about travelling to Paris. Such an individual could match the profile of a terrorist. Yet, he or she could also match the profile of a researcher seeking to understand ISIS as an organisation, its criminal networks and its reach into Western countries. Or the individual may simply have travelled to the Middle East for personal reasons, have an interest in collecting guns and wish to take a holiday in Paris, with no underlying motive linking these actions and interests together.
Big data analysis is meant to bring a measure of objectivity to intelligence-gathering, profiling possible suspects on the basis of clear criteria. In the process, analysts assume that they can eschew the subjective assumptions (even prejudices)