Vanishing Surveillance: Why Seeing What is Watching Us Matters
Prepared for the Office of the Privacy Commissioner of Canada by David Murakami Wood
DISCLAIMER: The opinions expressed in the reports and related summaries are those of the authors and do not necessarily reflect those of the Office of the Privacy Commissioner of Canada.
The twenty-first century is seeing a rapid transformation of our societies of into ubiquitous or pervasive surveillance societies. The phrase ubiquitous surveillance, which I often abbreviate to ‘ubisurv’ is adapted from the term ‘ubiquitous computing’ (or ‘ubicomp’), sometimes termed ‘pervasive computing’ or ‘ambient intelligence’ (AmI) and the adaption is not simply a language game. Ubisurv rests to a large extent on ubicomp. The gradual infiltration of computing and communications devices into increasingly mundane objects and structures and now often, living things, is one of the key technological enablers of ubiquitous surveillance. This development enables many other processes and activities too, and I would not want to imply that surveillance is the only or the inevitable use for such technologies, but it is a significant one. Nor would I argue that all the possible surveillant uses of ubicomp are simply by their existence or in their use, morally bad or suspect: surveillance underpins many activities that enable societies to function, good and bad.
There are many individual technological developments that one could chose to give as examples, from geodemographic and geolocation systems to the complex augmented reality environments created by artists. However, the ubiquity of surveillance is not simply a matter of there being ‘a lot of it about’, it is also a question of the way in which surveillance is at the same time less and less obvious even as it increases. Surveillance is vanishing. This poses particular challenges for both regulators and the public. We are used to there being covert surveillance conducted by particular organisations like the intelligence services like CSIS, and by police and private investigators, and we assume that such covert surveillance is closely regulated. In fact, such surveillance already raises many problems, but it is the vanishing of what we are used to thinking of as more overt and public surveillance that is the focus of this article.
For example, when we think of public area surveillance, we tend to conjure up images of video surveillance cameras (Closed Circuit Television or CCTV). We can see these cameras on their poles or mounted on walls and ceilings – they are obvious and visible. And in Canada, as well as in many other jurisdictions, we make them more visible still by requiring signage, informing users of the space who is responsible for the surveillance. In other words, regardless of our specific opinions on surveillance, we tend to link the visibility of the surveillance technologies to our ability to know we are being watched, to hold the operators of the systems to account through regulators and so on, and beyond that, to be able to make informed decisions about political and social action with regard to surveillance and privacy.
However, this era of visible surveillance technologies may well be coming to an end, and with it, potentially, our ability to know, to hold to account, to regulate and to make informed political decisions about surveillance, as well as the much forecast but still not quite upon us, end of privacy. If surveillance systems in general increasingly vanish, then we faced with the same kinds of problems as we face with covert surveillance with all forms of surveillance, used by a vastly expanded array of institutions and individuals than just those already problematic bodies like the intelligence services and police.
So how is surveillance vanishing? Though there are many ways, but for the sake of clarity, I identify three trends, each with one or two examples. The first is simply that surveillance technologies are getting smaller, whilst increasing in power and range of capabilities. The second is that surveillance technologies are more mobile, whilst remaining connected, and increasingly do not look like surveillance technologies, and may not even appear to be artificial. Finally, the accessibility and cost of sophisticated surveillance systems is decreasing all the time leading to what one could call a ‘democratization’ of surveillance, but putting powerful tools in a much wider variety of hands.
Binding all of these together is the ability to store and sort ever larger amounts of data on databases, whether these are the ‘server farms’ of private corporations like Google, Apple, Microsoft and Facebook as they gradually become the default option for personal data storage through cloud computing or the data warehouses of state organisations like the FBI’s Investigative Data Warehouse (IDW) which aims to suck in data on crime, criminals, as well as crime threats and potential criminals, from all over the world, including CCTV footage. The data analytic processes that take place to sort, sift and connect such data, from Internet traffic patterns to social networks to faces in the crowd, also take place largely invisibly and often beyond the boundaries of jurisdictions that may be concerned about the effects on their citizens.
1. From RFID Chips to Smart Dust and Beyond
We are already familiar with the use of Radio Frequency Identification (RFID). There are two types of RFID microchips or tags. Passive chips respond only to an active instrument or reader of some kind. Active RFID chips emit a continuous or periodic limited range radio signal that can be picked up by receivers. Until recently the use of RFID has been restricted to large shipping containers (ports being a major area of vulnerability to smuggling, illegal immigration and terrorist attack, but very difficult to police effectively by traditional methods), as well as consumer goods. Electronic tags have also become increasingly common in the UK and USA for those with judicial restrictions that tie them to a particular area (sometimes as limited as an individual house) for all or some hours of the day, which communicate either constantly or at regular intervals with a receiver, increasingly Global Positioning Systems (GPS) rather than simple radio monitoring.
However, recently, a notable change has occurred: the implantation of RFID chips into living beings. Racehorses and pets were the first groups to be targeted in this way. For pets, RFID chips containing information about immunisation records and ownership have gradually replaced quarantine requirements in the EU and elsewhere. For humans, the first use of RFID implants has been in elderly people suffering from degenerative diseases in the United States to enable carers to locate them easily and prevent them from wandering and possibly endangering themselves. Researchers and technological enthusiasts have also been implanting themselves with chips for several years now in order to be able to automatically performs small household tasks and at least one chain of Spanish nightclub briefly offered patrons the chance to have cash and access privileges held on implanted chipsFootnote 1.
There have been very few instances of workplace chipping. In February 2006 a security company on Ohio, USA, was reported to have implanted two of its workers with RFID chips to allow them to access company propertyFootnote 2. Although such an invasive procedure was carried out voluntarily, it raises enormous questions about the integrity of the body and privacy in relation to employers.
While small these RFID chips have up until recently been relatively visible. However, workable sensors are now being developed that function at ever-smaller scales. Dust Networks, a spin off company from Berkeley University in California has been producing 0.2mm2 chips since the early 2000s, and these devices are designed to work as distributed, wireless, multi-functional senor platforms for a variety of applications. The ‘motes’ made by Dust Networks are already starting to look large in comparison to the tiny RFID chips not being made by Hitachi of Japan, which has now demonstrated working examples at 0.025 mm2, the smallest working sensors currently known. It does not seem unlikely that such chips will continue to decrease in size.
2. Surveillance on the Move
The decreasing size of surveillance technologies has allowed surveillance to become increasingly portable and more covert. This is also enabled by wireless networking. However, the vanishing of surveillance is also encouraged by new types of surveillance platforms which can use vertical mobility to be more distant and able to ‘get away’ – the militaries of the USA, Israel, the United Kingdom and India, amongst many others, now rely heavily on Unmanned Aerial Vehicles (UAVs) for reconnaissance (as well as warfighting operations). However, UAVs are not restricted to the battlefield: the USA is now patrolling both the Mexican and Canadian borders with the same Predator drones it uses in Pakistan, and police in the UK, the USA and Italy have all purchased camera-equipped Micro-UAVs (MAVs) for urban surveillance operations.
Such MAVs, through their mobility, can bypass the laws on signage for video surveillance (although they have had problems with civil aviation regulations, at least in the UK). However they are still visible. Current developments may change this. The major area of research and development expansion in mobile surveillance devices is in biomimetic technologies. Biomimetics are machines that imitate naturally-occurring animals or plants. The most common biomimetic devices tend to be birds, snakes, and insects. AeroVironment of the USA recently demonstrated a working partially radio-controlled and partially autonomous robotic ‘Nano Hummingbird’Footnote 3. Technion, the Israeli Institute of Technology, has been developing a ‘snakebot’, a camera that moves like a snake, and could, they say, be used for search and rescue in collapsed buildings and so onFootnote 4. These devices are therefore not necessarily surveillance devices in themselves but as ‘platforms’ could be used for both surveillance and weapons deployment. That this is undoubtedly an aim is shown by the funding of such research: for example, AeroVironment is funded largely by the US Defense Advanced Research Projects Agency (DARPA).
The trend here is towards what paranoids have long believed – that anything could be a surveillance device. Clearly there are many technological obstacles to overcome, in the case of the Nano Hummingbird, the drawback as far as those who would want a truly covert device are concerned, is sound – it really does hum, and rather noticeably.
3. Everyone is a Spy
As the bleeding-edge the surveillance market produces ever smaller and more powerful surveillance devices, this further encourages the easy and cheaper availability of what were previously cutting-edge technologies. This has encouraged the routine use of ‘just-in-case’ self-protective surveillance, for example the ‘helmet cams’ employed by urban cyclists in the United Kingdom, which are often no different from the cameras worn by police to film interactions with the public. This use of personal surveillance technologies has been said by the Information Commissioner in the UK not to be covered by regulation in that jurisdictionFootnote 5 but it does pose significant privacy questions, especially as many cyclists post the videos they take of what they perceive to be dangerous (if not actually illegal) behaviour by drivers on publically accessible video-sharing sites like You Tube or Vimeo.
However, once again, these technologies will seem to be rather primitive in a few years’ time. Imaging devices have become more sophisticated. Cameras have been developed that can fir on the head of a pin, albeit with poor resolutionsFootnote 6. At border crossing points, we have seen the increasing deployment of body-scanners (or virtual strip searches), following a number of incidents of attempted terrorism using concealed explosives. So far such equipment is bulky and indeed is designed at least partly to create a theatrical atmosphere in which the passenger is made conscious of security and surveillance procedures. However, research is pointing in directions which indicate that the bulky visibility of body scanning may not be necessary. For terahertz wave technology, currently one of the least popular of the three main imaging technologies employed for body scanning, has now been shown to be able to work well, at least theoretically, in much smaller devicesFootnote 7. The output is of video standard and could be built into arrays using standard silicon CMOS technology, in other words, it would be something that could be added onto small, cheap and portable) equipment. The era of home cameras with body-scanning capability may not be far away.
What is to be Done?
There are many questions that arise from the simultaneous ubiquity and vanishing of surveillance. Some of these are social, or socio-technological – for example, how do these developments potentially affect the way we think about ourselves and our relationships? Social trust is the first casualty in surveillance societies, but ubiquitous and invisible surveillance might threaten to annihilate trust completely. And sophisticated covert home surveillance of friends and neighbours already appears to be a growing trend in North AmericaFootnote 8. On the other hand, one must consider whether such developments could be positive or even empowering. We have seen, for example that the widespread availability of small cameras has helped to make police and other public authorities and private organisations more accountable, as in the case of the death of Stephen Tomlinson at the G20 protests in London in 2008 or the complaints of police overreaction in Toronto in 2010. Technologies therefore develop not just in response to the designs of their inventors and vendors but also in response to human use in practice, however not all humans are equal and some uses will outweigh others in influence.
A key question concerns the political economy of vanishing surveillance. At present the research and development appears to remain strongly linked to military aims and objectives, or if not that, to the goals of a very small number of dominant private corporations. In either case, questions of design, purpose and accountability are significant ones, as well as the influence of particular market, rather than social, demands.
This then leads to the questions, which will most concern the Privacy Commissioner and the public: how can current and future law, and regulators, cope? How can both regulators and the general public respond to the vanishing of surveillance? It is already clear that politicians and the law are lagging behind the progress of technological development. There are few politicians with the scientific or technical knowledge to understand or assess developments in this area, and in addition, with the dominance of neo-liberal economic thinking, technological development is an area often seen as something that should be left entirely to the market, and in which government should involved itself as little as possible. This attitude unfortunately feeds back into the general technological ignorance of policy-makers. It also increases the opacity of the development process to the public and means that the politics of ubisurv increasingly vanishes behind the closed doors of producer groups, programmers, and specialist Internet webzines or books only likely to be read by the already technologically literate.
So what are the options? There are many but I will put forward five basic policy and regulatory trajectories here:
- Continue to try to enforce existing rules and rights. The Office of the Privacy Commissioner has to work in the context in which it finds itself. The OPC could interpret its remit in as maximal a way as possible to deal with new developments and technologies and practices that were not necessarily envisaged by the laws that empower it. This will, of course, always leave the OPC vulnerable to legal challenges and accusations of over-extending its remit, however this may be better than demanding new legislation that is over-specific. It would inevitably also have to ignore in practice the fact that any discovered violations would be likely the tip of the iceberg and indicative of much more underlying activity.
- Do nothing - or even encourage some or all of these developments. It may be that the democratization of surveillance will be empowering and allow citizens to hold government and corporations accountable themselves. Whilst the OPC should always be on guard for individual privacy, it might be argued that it should also encourage citizens to look after their own informational relationships with organizations as much as they are able to do so, and it should therefore embrace transparency and surveillance as tools of accountability. This would essentially mean that regulation would be applied to privacy in increasingly specific and circumscribed spatial locations and/or times, such as inside the home but not, for example on the streets. Essentially this would mean the end of privacy in any public or quasi-public setting.
- Encourage technological solutions, such as anti-surveillance and Privacy by Design and other interventions at the R&D stage. The latter solution has been the one advocated most notably by the Office of the Information and Privacy Commissioner of Ontario, and if technology is seem to be moving ahead of regulation, it has its attractions. However, it runs into several problems, most notably the need for close cooperation between regulators and developers, many of which have no particular interest in placating the regulators of relatively small economies. There are successful examples, most notably the camera ‘click’ sound required for all digital cameras by Korea and Japan. This was introduced in the wake if scandals about covert ‘upskirt’ photography and, with the centrality of the two countries in the manufacture of digital cameras, had consequently become the de facto standard for all digital cameras even before similar laws were introduced in the USA in 2009, for example. One can see how this might be applied in the case of cameras incorporating invasive new techniques like body-scanning or indeed, in requiring rather than eliminating the hum from the Nano Hummingbird in domestic markets, but could such a method be applied in the case of distributed network of sub-microscopic chips? What would the equivalent be in this case? Useful as it may be, Privacy by Design would seem not to address the question of vanishing surveillance per se.
- Market solutions. If data about identity, location and connection are valuable in an information society, surely anonymity, unlocatability and disconnection are also economically valuable. This being the case, an argument could be made for allowing the market to produce solutions to the problems posed by intrusive technologies. One can already see research into anti-surveillance devices for example, real-time video erasureFootnote 9 and light-bending materials that create a form of invisibilityFootnote 10. However demand in markets is not equally distributed amongst individuals but determined by existing distributions of wealth and power, and if market forces determined the practical direction of privacy rights, this would be likely to harm the already vulnerable and empower the already relatively powerful, extending existing divisions and inequalities.
- Hard Regulation. There may be a case in some instances for bans on the production and/or sale of some surveillance technologies that might be considered simply too much of a risk to privacy rights. One might, for example, adopt such a solution in the case of personal body-scanning cameras. The problem with such a solution in the case of vanishing surveillance is how one would know in the first place that some such technologies were being used, and therefore a major problem of detection and enforcement, which would require a major increase in specialist technology, personnel and funding for regulators, none of which would seem likely. It might also encourage a shadow market in banned surveillance technologies. Hard regulation does however have an attractive symbolic dimension.
It is more likely in practice that a mixture of approaches will be adopted. It may be that Privacy by Design is more appropriate in one area, hard regulation in another, and a more positive or laissez-faire approach in yet another. But one thing is certain: the job of safeguarding privacy is about to get a lot more difficult.
- Date modified: