Automatic Justice?: Technology, Crime, and Social Control
Abstract and Keywords
This chapter examines how forensic science and technology are reshaping crime investigation, prosecution, and the administration of criminal justice. It highlights the profound effect of new scientific techniques, data collection devices, and mathematical analysis on the traditional criminal justice system. These blur procedural boundaries that have hitherto been central, while automating and procedurally compressing the entire criminal justice process. Technological innovation has also resulted in mass surveillance and eroded ‘double jeopardy’ protections due to scientific advances that enable the revisiting of conclusions reached long ago. These innovations point towards a system of ‘automatic justice’ that minimizes human agency and undercut traditional due process safeguards that have hitherto been central to the criminal justice model. To rebalance the relationship between state and citizen in a system of automatic criminal justice, we may need to accept the limitations of the existing criminal procedure framework and deploy privacy and data protection law.
*Technological and scientific developments have profound implications for the traditional paradigm of criminal justice and for the established relationship between state and citizen. Advances in the sophistication, range and ubiquity of forensic technology are part and parcel of ‘the creeping scientization of factual inquiry’ predicted nearly 20 years ago by Mirjan Damaška (1997: 143). The goal of this chapter is to examine the implications for the traditional model of criminal justice of new technologies in the investigation and prosecution of crime (Bowling, Marks, and Murphy 2008). Our argument is that we are moving rapidly towards an increasingly automated justice system that undercuts the safeguards of the traditional criminal justice model. This system favours efficiency and effectiveness over traditional due process safeguards and is taking on a life of its own as it becomes increasingly mediated by certain types of technology that minimize human agency. This, we argue, is creating a system of ‘automatic justice’. In order to rebalance the relationship between state and citizen in such a system, we may need to accept the limitations of (p. 706) the existing criminal procedure framework and seek instead to unpick its core values for more effective deployment in the fields of privacy and data protection law, themselves rapidly evolving areas of law of increasing relevance to criminal justice.
The chapter proceeds as follows: first, we provide an outline sketch of the traditional paradigm of criminal justice and its significance for the relationship between state and citizen; second, we explain contemporary trends in criminal justice and explore how shifts in emphasis within the criminal justice system are facilitated and accelerated by advances in science and technology; third, we examine how technological innovations affect the internal and external stability of the traditional criminal justice model and the implications of these changes in terms of its fundamental values. We illustrate how the pressure of technological innovations has made the external and internal boundaries of the criminal justice model more porous even while becoming increasingly securitized. In the absence of any clear boundary between the criminal justice system and the outside world, the privacy and liberty of all citizens is vulnerable to arbitrary state intrusion. Fourth, and finally, we attempt to address the challenges identified by first exploring the possibility of applying criminal principles more broadly before briefly turning to privacy and data protection law which may be capable of offering an alternative and additional architecture for upholding similar values to those enshrined within it (Damaška 1997: 147).
2. The Traditional Criminal Justice Paradigm
The criminal law has traditionally embodied the most coercive powers of the state for regulating the behaviour of citizens. Although the traditional model of criminal justice is conceptually elusive, certain normative values can be identified. Herbert Packer, in his classic text, describes what he calls the ‘due process’ model as containing an ‘obstacle course’ of procedural rules which safeguard against injustice while facilitating the pursuit of truth (1968: 163). In this sense, we can speak of an ‘inner morality’ to criminal procedure (Duff and others 2007: 51); the criminal trial is its focal point, its main event in pursuit of legitimate dispute resolution between state and citizen. A core value of any liberal democracy is the principle of minimum state intrusion. Agents of the state are granted powers to coerce citizens, to intrude into their private lives and to deprive them of their liberty—as a means to pursue the ends of community safety and public order. The corollary to the granting of these intrusive and coercive powers to the state is that they must only be used when they are lawful, necessary, and proportionate.
(p. 707) The criminal law consists of offences, the commission of which enables the state to punish a citizen by stigmatizing them, depriving them of their liberty or inflicting some other pains upon them. At the outset, and preceding the trial stage, criminal investigation necessitates reasonable, individualized, and articulated suspicion as a prerequisite for investigation; where there is sufficient evidence against a suspect, that individual is charged with a specific criminal offence and the individual’s legal status becomes that of a defendant. The trial is central in this model: investigative powers are predicated on their ability to serve the trial process by producing evidence that can be examined and tested in a court of law, the factual determination and fairness of which will determine whether punishment is justified. A fair trial, then, is a prerequisite of punishment and must take place within a reasonable time period. In the traditional model, the defendant either exits the criminal justice system at the conclusion of the process, or where punishment is justified, acquires the status of convict, and is punished by being retained within the system for a proportionate period of time.
The due process safeguards sewn into the traditional model have three overarching aims: (i) the minimization of state intrusion into the lives of citizens, (ii) the protection of human dignity, and (iii) upholding the legitimacy of state coercion and factual accuracy. Procedural propriety, concerned as it is with human dignity, is designed to ‘treat the accused as thinking, feeling, human subjects of official concern and respect, who are entitled to be given the opportunity to play an active part in procedures with a direct and possibly catastrophic impact on their welfare, rather than as objects of state control to be manipulated for the greater good (or some less worthy objective)’ (Roberts and Zuckerman 2010: 21). It is with this model in mind that we turn to modes of governance, trends and attendant technologies that are eroding and reshaping the criminal process.
3. Prevailing Modes of Governance and Paradigm Shifts in Criminal Justice
The criminal justice system has been subject to a great deal of change in the course of the last 40 years, often in complex and incoherent ways (O’Malley 1999). However, certain patterns and trends in criminal justice policy have tended to reflect prevailing modes of governance. A panoply of expressions have been proffered to capture the essence of the current mode of governance prevalent in Western liberal democracies: the hollowing out of the state (Rhodes 1994); the retreat of the state (Strange (p. 708) 1996); governing at a distance (Rose and Miller 1992); a state that steers rather than rows (Osborne and Gaebler 1992); ‘the facilitating state, the state as partner and animator rather than provider and manager’ (Rose 2000: 327). All reflect an ideology, neo-liberalism, in which the state seeks to enshrine the imperatives of economy, efficiency and effectiveness in its endeavours through the privatization of many state functions while embedding business model ideals in others. This comes in the form of the new public managerialism, which injects a commercial ethos into public provision. The reification of the market and its mechanisms and the intense auditing of public bodies are central tenets of this mode of governance (Jones 1993). As such, the language and ethos of commerce have spread throughout the criminal justice system. Police forces are now required to ‘do more with less’ in the face of increasingly austere budgetary constraints. The unwieldy edifice of the traditional ‘due process’ model of the criminal justice system, a system which prioritizes fairness over efficiency, is deemed unfit for purpose. Increasingly, aided by advances in technology and science and in the interests of efficiency, individuals are investigated, judged, and punished en masse and at a distance. Co-opting Rose and Miller’s ‘governing at a distance’, commentators now speak of a contemporary ‘punishment at a distance’ (Garland 2001: 179), a term which envisages a criminal justice system which regards individuals as little more than words and letters in a database.
We are witnessing a gradual movement away from the traditional, retrospective, individualized model of criminal justice, which prioritizes a deliberated and personalized approach to pursuing justice and truth, towards a prospective, aggregated model, which involves a more ostensibly efficient, yet impersonal and distanced, approach. ‘Actuarial justice’ is based on a ‘risk management’ or ‘actuarial’ approach to the regulation of crime and the administration of justice (Feeley and Simon 1994). Feeley and Simon have characterized this as a movement away from concern for the individual offender, towards an emphasis on aggregates and ‘the development of more cost-effective forms of custody and control and … new technologies to identify and classify risk’ (1992: 457). As described by Garland, ‘the individual is viewed not as a distinct, unique person, to be studied in depth and known by his or her peculiarities, but rather as a point plotted on an actuarial table’ (Garland 1997: 182). A central precept of actuarial justice, therefore, is that the system should be less concerned with traditional punishment based on downstream or after-the-fact goals such as retribution and rehabilitation. It should instead manage the risk presented by the dangerous and disorderly, using upstream or pre-emptive techniques of disruption, control, and containment. The shift from retribution and rehabilitation towards prevention means that the state seeks to identify potential criminals before they commit offences. In light of such trends, Malcolm Feeley predicted the eventual emergence of a ‘unified actuarial “system” that will completely transform the criminal process into an administrative system’ (2006: 231).
Actuarial justice takes many and varied forms and is closely related to ‘intelligence-led policing’. This is a future-oriented mode of policing in which crime data are (p. 709) collected, risks are assessed and policing strategies are formulated accordingly (Maguire 2000). This is a departure from the reactive methods of policing prevalent up to the 1970s when ‘[t]he main organizational requirement was to arrive, do something, and leave as quickly as possible’ (Sherman 2013: 378). Intelligence-led policing aims not merely to detect, investigate, and prosecute offences, but to deter and disrupt the activities of those deemed likely to commit crime in the future. This form of policing has included the improper use of stop-and-search powers to deter and control certain types of behaviour, rather than to allay reasonable suspicions of criminal activity (Young 2008), heightened surveillance of ‘high risk’ residential areas (Joh 2014), the use of no-fly and other such watch list and blacklist regimes, citizenship-stripping and the use of newly created disposal alternatives to criminal prosecution in the form of civil preventive orders (such as antisocial behaviour orders (ASBOs) and Terrorism Prevention and Investigation Measures (TPIMs)) all of which result in often stigmatic and punitive repercussions for the individual involved, while obviating the procedural safeguards of the criminal trial. Such alternatives to criminalization ‘define status, impose surveillance, and enforce obligations designed variously to control, restrict or exclude’ (Zedner 2010: 396).
Surveillance is a necessary correlative of risk-based actuarial criminal justice (Lyon 2003: 8). While surveillance may be deployed in a wide-ranging variety of contexts and for similarly myriad purposes, in the context of crime control its purpose is to distil knowledge from uncertainty. Accordingly, it can be claimed that ‘[t]he yearning for security drives the insatiable quest for more and better knowledge of risk’ (Ericson and Haggerty 1997: 85). Michel Foucault’s panopticon (1977) (as well as Orwell’s Big Brother (1949)), conceived and abstracted in its broadest sense as unidirectional surveillance emanating from a monolithic, bureaucratic surveillance state has been a dominant and prevailing theoretical model of surveillance (Haggerty and Ericson 2000: 606). It is a model, however, that is indicative of the post-war state, cumbersome and lacking in dynamism and fluidity, which differs considerably from late modern surveillance, which traverses borders and institutions, both public and private, and expands in disparate and varied forms. At the turn of the century, Haggerty and Ericson, drawing on the influential work of Gilles Deleuze and Felix Guattari (1988), conceived of a ‘surveillant assemblage’, which expands upon, rather than departs from, the panoptic surveillance model (2000). This surveillant assemblage is rhizomatic in that surveillant capabilities develop and expand in a vast number of ways and in different contexts, combining to provide a complementary surveillant visage. One can also speak of different assemblages which combine and plug into each other (Haggerty and Ericson 2000: 608).
A further change in the criminal justice landscape that forms part of the backdrop to our discussion is the idea of ‘simple, speedy, summary justice’ (Home Office, Department of Constitutional Affairs and Attorney General’s Office 2006). Referred to as a ‘new form of administrative justice’ (Jackson 2008: 271), it overlaps with actuarial justice to the extent that both seek to divert potential offenders from the (p. 710) full rigour of criminal proceedings with alternative disposal proceedings. Simple, speedy, summary justice seeks to do this by increasing the range and uptake of pretrial disposal procedures and its stated aim is the saving of expenditure: ‘What is changing is the scale of the number of cases proposed for diversion from the courts and the punitive steps that may be taken by prosecutors against those offenders who admit their guilt’ (Jackson 2008: 271). New forms of disposal include conditional cautions, reprimands, warnings for the young, fixed penalty notices, and a strengthening of administrative and financial incentives to admit offences and plead guilty early in the criminal justice process. Technology is shaped by these organizational goals, themselves moulded by prevailing modes of governance. In the case of the criminal process, incentives to ensure economy, efficiency and effectiveness embed administrative criminal processes that bypass fairness and legitimacy. The imperative of efficiency acts as an institutional incentive to adopt certain forms of technology and so ‘[n]ew technologies are routinely sold to criminal justice practitioners on their promised efficiency gains and cost savings’ (Haggerty 2004: 493).
4. ‘New Surveillance’, ‘New Forensics’, and ‘Big Data’
Three overlapping terms identify and capture what is novel and significant about the use of technology in the criminal justice field today: ‘new surveillance’, ‘second generation’ forensics, and ‘big data’. Although each term could be used to describe applications of the same particular technologies—such as DNA profiling and CCTV—each term connotes its own implications for the criminal justice context. ‘New surveillance’ involves ‘scrutiny through the use of technical means to extract or create personal or group data, whether from individuals or contexts’ (Marx 2005). The term is used to convey the relative intrusiveness of technology in extending the senses and rendering ‘visible much that was previously imperceptible’ (Kerr and McGill 2007: 393) and enabling law enforcement to undercut procedural safeguards by obtaining more information about citizens than would be available from a traditional search or questioning (Marx 2005). Erin Murphy (2007: 728–729) has written in detail on the characteristics of what she calls ‘second generation’ forensics, of which the following are pertinent to this chapter:
(i) the highly specialized knowledge and expertise underlying them—which makes their workings less accessible and transparent to laypersons than traditional forensics;
(p. 711) (ii) the sophistication of the underlying science—which enables them to at least be portrayed as providing proof to a degree of scientific certainty and capable of providing conclusive proof of guilt. Whereas traditional forensics have generally played a supporting role (to eyewitness testimony and confession evidence, for example) second generation technologies are more frequently deployed as the sole piece of evidence; and
(iii) their dependence on databases and their ability to reveal a broad range of information (as opposed to being confined to the confirmation or denial of a specific question such as an identification) and their consequently deeper intrusion into privacy than traditional forensics.
‘Big data’ is generally accepted as an abbreviated expression for the application of artificial intelligence to the vast amount of digitized data now available and in this context much of the data will be obtained from ‘new surveillance’ and ‘second generation’ forensics. Such new methods of intelligence gathering are used to obtain more information than could be obtained from the traditional investigative techniques of questioning a suspect or subjecting them or their houses to a physical search, and facilitate the automatic analysis of the data inputted. As pithily summarized by Elizabeth E. Joh, ‘the age of “big data” has come to policing’ (2014: 35).
The combination of ubiquitous digital records and computer processing power has revolutionized profiling and social network analysis.1 Every time we make a phone call, send an email, browse the Internet, or walk down the high street, our actions may be monitored and recorded; the collection and processing of personal data has become pervasive and routine (House of Lords Select Committee on the Constitution 2009: 5). Dataveillance—‘the monitoring of citizens on the basis of their online data’ (Van Dijck 2014: 205)—is a paradigm example of new surveillance. Advances in the use of mathematically based analytical tools to detect patterns in large sets of data have facilitated profiling (a method of data mining) (Bosco and others 2015: 7) ‘whereby a set of characteristics of a particular class of person is inferred from past experience, and data-holdings are then searched for individuals with a close fit to that set of characteristics’ (Clarke 1993: 405), in order to establish and identify patterns of suspicious behaviour. There is a consensus among European data protection authorities (DPAs) that the three principal characteristics of profiling are (1) that it is based on collection, storage and/or analysis of different kinds of data and (2) on automated processing and electronic means and (3) with an objective of prediction or analysis of personal aspects or personality and creation of a profile. Additionally, a fourth key principle for some DPAs is that the profiling results in legal consequences for the data subject (Bosco and others 2015: 23). Data mining tools are used by the police to identify those who should be subjected to the growing proliferation of alternative control measures to the trial process outlined in section 2. (p. 712) New surveillance coupled with risk-based actuarial techniques and data mining technology seeks to make the best use of limited resources, and contributes to the increasing contemporary reliance on the intelligence-gathering, or absorbent, function of policing (Brodeur 1983: 513). Technology facilitates the transmission of information through space (communication) and through time (storage) (Hilbert 2012: 9). There is no physical barrier to storing all data—such as CCTV footage, facial images—indefinitely. Increasingly sophisticated means of recognition technologies and search tools will ‘one day make it possible to “Google spacetime”, to find the location of a specified individual on any particular time and date’ (Royal Academy of Engineering 2007: 7).
Unlike traditional forensic science techniques such as ink fingerprinting and handwriting analysis, ‘second generation’ forensic technologies (Murphy 2007) such as digital fingerprinting and DNA typing draw on vast digital databases and do not require the police to first identify a suspect. Whereas first-generation techniques were mainly used to confirm or deny suspicion, second-generation techniques have heightened investigative capacities. The evolution of fingerprint technology illustrates this trajectory. Once limited to the identification of fingerprint patterns in the search of individualized matches, the addition of mass spectrometry to fingerprint analysis has enabled detailed examination of these marks to reveal quite a lot about the personal and behavioural characteristics of the person who left the trace: their likely gender, the hygiene products they have used, the food ingested, whether or not they are a smoker, etc. As Alex Iszatt, writing in the Police Oracle, explained: ‘By analysing the constituent parts of the finger impression, we can profile the habits of the offender’ (2014). This development means that a print examination that may have failed to produce a suspect in the past can, if combined with data-mining tools, produce a larger pool of potential suspects than traditional fingerprint identification. It also makes a database of the personal habits and lifestyle of the population a potentially useful resource for crime investigators, thereby providing further fodder for those seeking to legitimize the collection of personal data en masse and in the absence of prior suspicion. Fingerprint technology now enables digital fingerprints to be easily collected and immediately compared on the spot with others contained in a database.
‘Big data’ analysis, ‘new surveillance’, and ‘second generation’ forensic sciences lend themselves more easily to the new ‘actuarial model’ of criminal justice—based on computerized risk prediction and purportedly objective and conclusive results—than to individualized criminalization. The streamlined forensic reporting process recently introduced in English courts has the stated aim of cutting ‘costs and speed[ing] up the production of a forensic result: to undertake just enough forensic work to support a charge or secure an early guilty plea’ (Forensic Access 2014). (p. 713) New technologies play a key role in accelerating the trend towards ‘simple, speedy, summary justice’ by encouraging and facilitating diversion and discouraging costly challenges to prosecution evidence.
This decade a whole raft of miscarriages of justice has come to light in which wrongful convictions are attributed to mistaken expert opinion or the ineptitude of the traditional criminal justice model (populated as it is by non-scientific personnel) to correctly interpret technical and scientific evidence. Several high-profile reports have been published in common law jurisdictions around the world seeking to address the problems.2 The perceived crisis in the traditional criminal justice model has deepened distrust of subjective knowledge and this has now extended from distrust in the layperson to distrust in expert opinion. The push is now to more and more purportedly objective data (Fenton 2013). Methods of decision making—understood to be objective, mathematical, and scientific—are signalled as less biased than the ‘common sense’ exercised by human decision makers (Palmer 2012) and more adept at dealing with both the complexity and amount of scientific data:
[A]s the gulf widens between reality as perceived by our natural sensory apparatus and reality as revealed by prosthetic devices designed to discover the world beyond the reach of this apparatus, the importance of the human sense for factual inquiries has begun to decline.
(Damaška 1997: 143)
Forensic investigations are themselves increasingly automated (so-called ‘push-button forensics’) and this results in loss of understanding in the underlying concepts of the investigation among not only the recipients of the information (law enforcement agents, defendants, and courts) but also the scientists actually conducting the forensic investigation (James and Gladyshev 2013). Mobile handheld devices enable DNA and fingerprints to be taken from a suspect and analysed on the spot, providing an immediate—albeit potentially provisional—conclusion, and persuasive accusation. Such is the apparent superiority of automated methods to human decision-making that an increasing number of courts are aided by ‘safety assessment tools’ to reach decisions on bail applications and sentences (Leopold 2014; Dewan 2015). The reputed success of these tools calls into question the value of human decision-making with its attendant biases and lack of technical comprehension. Psychologist and academic, Mandeep Dhami doesn’t believe that the decision-makers in tomorrow’s courtroom will be human. Her research concludes that magistrates are ‘not as good as computers’ at reaching decisions and claims it is conceivable that magistrates ‘could be replaced by automated systems’ (Dhami and Ayton 2001: 163). Other researchers suggest that data driven predictions and automated surveillance may actually reinforce prejudices and may even introduce a less readily transparent and ‘unintentional, embedded prejudice’ of their own (Macnish 2012: 151).
(p. 714) 5. Automatic Criminal Justice?
Reliance on databases ‘that locate, identify, register, record, classify and validate or generate grounds for suspicion’ (Marx 2007: 47) results in ‘a broadening from the traditional targeting of a specific suspect, to categorical suspicion’ (Marx 2001). The ‘net-widening’ of those subjected to surveillance and control in the intelligence-led model of policing is facilitated and accelerated by the present-day ubiquity of technologies of the so-called ‘new surveillance’. A common criticism of mass surveillance and data retention is that it makes ‘all citizens suspects’ and this is frequently deemed to be objectionable in and of itself.3 A system of crime control has emerged that operates in parallel to the traditional criminal justice system. The parallel system treats all citizens as suspicious and its surveillance is not predicated on individualized suspicion but is ubiquitous. It metes out heightened surveillance, control measures and punishments and places citizens on blacklists—sometimes on the basis of secret intelligence, without even communicating with the subject, in the absence of judicial oversight, and without providing any mechanism for redress (Sullivan and Hayes 2011).
Legislative changes have already made major inroads on the principle of double jeopardy in recent years in order to take advantage of scientific advances.4 A steady stream of scientific and technological advances will continue to increase the likelihood that new and compelling evidence that was unknown at the time of a defendant’s acquittal will be discovered. Such developments blur the boundary between the absolved and the suspect and undermine the finality inherent in the traditional model. The parallel system challenges the external boundaries of the criminal justice system at both the exit and entry points; bypassing the procedural framework of the criminal justice system and the distinctions between citizen, suspect, defendant, convict, and acquitted.
We are witnessing a simultaneous dissolution of the procedural infrastructure within the criminal justice system. The result of the actuarial trends described above, fortified by technological and scientific innovations, is that the traditional law enforcement regime has been ‘transformed from an investigatory into an accusatory and dispository regime’ (Jackson 2008: 272). Where wrongdoing is detected and proven, justice can be meted out instantly through the use of conditional cautions, fixed penalties for disorder, etc. rather like parking or speeding tickets. Aided by technology, the investigative and probative stages of the criminal justice process—which have traditionally been kept quite separate—are now merging into a single process that may be almost devoid of human judgement or engagement.
Automation is now routine in some spheres of policing and punishment. Take, for example, the investigation of motor vehicle licence and insurance investigation. (p. 715) In the UK, is has been estimated that the police seize about 500 vehicles per day where a police officer has reasonable grounds to believe that it is being driven without insurance (McGarry 2011: 220; Motor Insurers’ Bureau 2010). This intervention can arise from a routine police check, or a specialized ‘crackdown’ operation that are now used extensively by police forces across the United Kingdom. The process involves the use of Automatic Number Plate Recognition (ANPR) checks on vehicles that are then cross-checked against the Police National Computer (which is linked to the Motor Insurance Database (MID) and the Driver and Vehicle Licensing Agency (DVLA)) to determine whether there is a record of insurance (Kinsella and McGarry 2011). If there is not, a large sticker is placed on the windscreen stating that the vehicle has been ‘seized by the police driven without insurance’ and the vehicle is immediately loaded onto a car transporter and removed to the car pound. With a high degree of automation, the offence is surveilled, investigated, detected, and proven and the ‘offender’ punished, named, and shamed. Connecting the dots, the near future of road policing is a system in which cameras connected to computers read motor vehicle licence plates automatically, recognize the faces of the driver and front seat passenger, detect the speed of the vehicle, connect these data to licensing and criminal databases, issue penalties and deploy officers to demobilize a vehicle and remove it from the road. In the context of the idea of the ‘internet of things’—wherein mechanical devices speak directly to each other through computer networks—it is possible to imagine a world in which citizens are investigated, evidence collected against them, a judgment of guilt reached and a penalty issued without the participation of a human being at any stage.
There are good reasons to anticipate that ‘automatic justice’ will soon be prominent in ordinary street policing. As discussed above, pre-emptive encounters inspired by technologies of risk are now permeating police activity at all levels (Hudson 2003). New devices—such as body-worn video cameras and those installed in vehicles—enable police officers to record encounters with the public in unprecedented ways. The evidence collected through these devices can be compelling and instantly reviewed by both police and suspect. The rolling out of tailored computer tablets will provide instant access to all police and other databases—both national and transnational—on all matters from criminal records of names, faces, fingerprints, DNA as well as criminal intelligence, licensing, insurance, welfare benefits, and other personal information. The vision of the leadership of the police is to ‘use digital to connect the criminal justice system from the very first report of a crime through to a court appearance, an end-to-end service’ (Hogan-Howe 2015). The adoption of technology within law enforcement is perceived as ‘reaping the benefits of digital technology and smarter ways of working’ including an improvement in ‘quality of service’, productivity, efficiency, and effectiveness (Hogan-Howe 2015). The Metropolitan Police Commissioner’s drive towards a ‘truly digital police force’ points in the direction of automatic justice.
(p. 716) 6. Some Implications for Criminal Justice Values
Twenty years ago Mirjan Damaška suggested that:
[T]he Anglo-American procedural environment is poorly adapted to the use of scientific information … the scientization of proof is likely to exacerbate the presently minor frictions within traditional procedural arrangements. Their further deterioration should be considered likely on this ground alone.
The developments outlined above strain the traditional concept of the trial as ‘a climactic event’ and beg the question of whether the public trial retains its status as the focal point in criminal procedure (Damaška 1997: 145). In facilitating and accelerating dramatic changes to the architecture of the criminal justice system, these new technological developments raise urgent questions of procedural legitimacy. For example: how and at what stage will the reliability and accuracy of evidence be challenged? What is the fate of due process safeguards such as the presumption of innocence, the right to silence, the requirement of reasonable suspicion for the exercise of police powers, the right to a trial within a reasonable time, the principle of equality of arms and the right of confrontation? All are expressed in the procedural components of the traditional criminal justice model, but are rendered problematic in an automatic criminal justice system.
Unfortunately, as McGarry points out, automatic systems of policing—such as the investigation of uninsured driving—reverse the burden of proof, sometimes make errors, are experienced as heavy-handed, have few safeguards or means of correcting mistakes, and have a significant impact on innocent individuals punished and shamed in this way (McGarry 2011). The challenges posed for due process by the trends and technologies outlined above are summarized by Wigan and Clarke in an article about the unintended consequences of ‘Big Data’:
Decision making comes to be based on difficult-to-comprehend and low-quality data that is nonetheless treated as authoritative. Consequences include unclear accusations, unknown accusers, inversion of the onus of proof, and hence denial of due process. Further extension includes ex-ante discrimination and guilty prediction, and a prevailing climate of suspicion. Franz Kafka could paint the picture, but could not foresee the specifics of the enabling technology.
Implicit in this paragraph and in the criticisms made by criminal law scholars of ‘speedy justice’ is the failure to engage with the suspect as an agent and the risk of persons being treated as guilty on the basis of inaccurate and unchallenged evidence (Jackson 2008).
Thus far, data mining and profiling are not reliably accurate in their behavioural predictions. As observed by Daniel Solove, given that most citizens are subjected (p. 717) to data-mining techniques, even a very small false positive rate (1%) will result in a large number of innocent people being flagged as suspicious (Solove 2008: 353). As Valsamis Mitsilegas has pointed out, many will not have the possibility of knowing, let alone contesting, such assessments (where they result in the flagging of suspicious behaviour for example) (2014). Even where they do, the assumptions that algorithms contain may not only be incorrect, but will also be difficult to spot because they are ‘hidden in the architecture of computer code’ (Solove 2008: 358). Legal practitioners have expressed concern about the ‘aura of infallibility’ that surrounds mathematically generated information, deterring attempts to understand the process by which results are reached and inhibiting challenges to their accuracy (Eckes 2009: 31).
High levels of accuracy and certainty are required for criminal convictions (the case must be proven beyond reasonable doubt) on account of the cost of error both to the state (in terms of legitimacy and financial cost) and to the individual (in terms of stigma, financial loss, or deprivation of liberty) (Roberts and Zuckerman 2010: 17). The issue that needs to be considered in relation to profiling of suspicious behaviour is working out the cost of error to both the state and the individual. Such evaluations should consider not only the cost of being singled out for extra investigation or for inclusion on a blacklist and the service denials entailed (Ramsey 2014), but also the more abstract implications for rights such as the freedom of association and the right to privacy, given that much of the data used in generating the profile will come from the lawful exercise of such rights (Solove 2008: 358).
Law enforcement agents may not always appreciate the potential for error in new technologies and defer their judgment and decision-making to technology. The relevance (Marks 2013) and reliability (Law Commission 2011) of scientific evidence are often incorrectly evaluated by legal systems. Concerns have been expressed about the so-called ‘CSI effect’ in relation to misplaced police, judicial and jury deference to forensic science. Scholars have been equally concerned with vulnerability of legitimacy of the verdict if courts are seen to defer to expertise: ‘The fear is spreading that courts are covertly delegating decision-making powers to an outsider [expert] without political legitimacy. Is the court’s nominal servant becoming its hidden master?’ (Damaška 1997: 151). According to Roberts and Zuckerman, if this fear regarding deference to science and technology turns out to be well-grounded, it would amount to a radical realignment of criminal trial procedure (2010: 489). Total reliance upon science represents ‘an abdication of responsibility by law’ (Alldridge 2000: 144).
It is clear that the criminal justice system will command more support from the public if its procedures, techniques, and outcomes are easily and widely comprehensible. This is undermined by a system of increasingly automatic justice that relies heavily on scientific procedures, mathematical calculations, and technical devices. The prospect of uniform data collection and permanent data retention has several implications for the criminal justice model. The criminal justice system is designed to not broadcast information about minor offences, and in many countries criminal (p. 718) records systems actively omit old offences when performing criminal records checks. The system thus favours forgiveness and rehabilitation, rather than permanently labelling people as criminals. Indiscriminate data retention conflicts with such constructive aims (Wigan and Clarke 2013: 51).
7. Meeting the Challenge
The challenge is how to ensure that ‘automatic justice’ retains the characteristics of ‘justice’ and is in accordance with fair trial rights. On the one hand, there are the challenges to criminal justice values within the traditional model, and on the other there are those posed by the emergence of the parallel model of crime control that may entail greater state intrusion and coercion than that of the traditional criminal law. Predictive mass surveillance (as opposed to targeted reactive surveillance) is largely unregulated by criminal justice procedures and yet it is arguably more intrusive on privacy rights on account of its endurance where data is stored indefinitely. The legality of much of it is also far from clear. Judgments of the European Court of Human Rights and the European Court of Justice have made clear that the mere retention of personal data, regardless of what is consequently done with it, amounts to intrusion into privacy (S and Marper v United Kingdom 2009: para 67). The court has also decided that blacklisting regimes must comply with fundamental rights including the right to a defence as guaranteed by Article 6 of the European Convention on Human Rights (Yassin Abdullah Kadi and Al Barakaat International Foundation v Council and Commission 2008). Mirelle Hildebrandt has described the fundamental principles of criminal procedure as historical artefacts that should not be taken for granted. They should instead be acknowledged as the foundations upon which any liberal democracy—including one operated by sophisticated technology—must rest (Hildebrandt 2008).5 We perceive three means of addressing the challenges posed by automatic justice to the fundamental values traditionally enshrined in criminal procedure: (i) extending the procedural requirements outside of the traditional criminal justice model; (ii) incorporating criminal justice values within the rapidly evolving and heavily contested field of data protection law; (iii) incorporating lessons learned in data protection law within the criminal justice system.
7.1 Extending the Scope of the Criminal Justice Model
In light of the range of alternative disposal measures and the wide reach of new technologies it is arguable that the threat of criminal investigation, prosecution and (p. 719) conviction no longer represents the most coercive arm of the state and that the citizen/suspect and suspect/convict distinctions no longer hold water. As has been persuasively argued by Liz Campbell, ‘this may imply that some of the traditional protections that relate to the criminal trial, strictly speaking, are valuable or necessary in a wider context’ (2013: 689). Because surveillant, investigative, probative, and punitive powers have migrated out of the criminal justice system and are now more widely diffused throughout society, there is a need for the due process protections with which they are associated to follow suit. Several scholars and courts have recognized the importance of extending judicial oversight and due process safeguards to the administration of potentially punitive measures such as watch lists (Ramsey 2014).
The ubiquity of new surveillance raises the question of what it means to be a suspect. Some scholars have gone so far as to claim ‘the concept of suspect is now thoroughly transformed, for we are no longer able to confine it to its juridical sense’ (Bauman and others 2014: 138). Liz Campbell argues that we should recapture the concept of suspect within its juridical sense by widening it to include an ‘interim categorisation of suspicion’ (Campbell 2010: 898) for those treated by the state in a manner that suggests they are not ‘wholly innocent’ (S and Marper v United Kingdom 2009: 89). This interim category is reserved for retaining information on a mass scale of certain targeted categories of people, premised on the perceived risk they pose. A status of ‘proto-suspect’ could entail a measure of suspect’s rights. As well as entailing an expansive interpretation of the presumption of innocence this reconceptualization of the suspect suggests the core of being a suspect is to be treated differently to others. There is a danger of conflating the concept of suspect with the concept of convict here however, which itself reflects a tension in the application of the presumption of innocence and the extent to which it should apply to pretrial measures as well as to the trial, let alone to the more recent debate over extending it to alternative disposal measures.
The apparent similarities between the treatment of convicts and the treatment of persons who are not even suspects gained ground for the right to privacy in S and Marper v United Kingdom (2009) by curtailing the retention of the DNA of persons other than those convicted of criminal offences. But the logical conclusion of applying the protection of the presumption of innocence only where persons are treated differently to those who are ‘wholly innocent’ fails to protect privacy in the long term if citizens are all subject to different forms of mass surveillance and data retention. In addition to further exploring the values underpinning the presumption of innocence—particularly in relation to its potential application to regulatory measures that look like criminal punishments (punishment as ‘communicative act’), it may be useful to think about what negative consequences, other than stigmatization, could be said to be at the core of being a suspect.6 We might look at the values underpinning the principle of double jeopardy for further guidance.7 An important value said to underpin the principle of double jeopardy is finality. There is value to the suspect and to society as a whole in accepting that a contested issue (i.e. the (p. 720) criminality of the suspect) has been resolved (Ashworth and Redmayne 2010: 399). In the absence of resolution one remains a suspect under permanent threat of state coercion into an indeterminate future. There are undoubtedly other important aspects to the status of suspect and this could prove useful terrain for those seeking to address the perceived injustices of mass surveillance.
7.2 Incorporating Developments in Privacy and Data Protection Law within the Criminal Justice System
Alldridge and Brants (2001: 5) argue that ‘as proactive policing increases in importance, consequently, so should attention be devoted to the claims of privacy in criminal procedure’.8 Evidence scholars have highlighted the threat posed to privacy by disproportionate intrusions in the form of the combined relaxation of the rules on the admissibility of bad character evidence, broad definition of such, and the wealth of personal data now accessible online. As Roberts and Zuckerman (2010: 599) put it, ‘empowering the authorities to require the accused to answer at large for his entire existence and moral character would not be compatible with liberal conceptions of limited government.’ Greater attention needs to be paid to protecting a defendant’s right to privacy within the traditional criminal justice model on account of the reach of ‘new surveillance’ and ‘second generation’ forensics.
We focus on the right to privacy on the basis that ‘privacy is to proactive policing what due process is to reactive policing’ (Alldridge and Brants 2001: 21). The raison d’être of laws on police searches is to protect a citizen’s personal space from arbitrary interference but new technologies bypass such safeguards by obviating the need for traditional stop and searches (R v Dyment 1988). New surveillance practices and technologies—such as thermal imaging and millimetre waves can investigate a person’s belongings without making physical contact and without the subject even being aware that they are being ‘searched’. The use of metadata from personal communications ‘may make it possible to create a both faithful and exhaustive map of a large portion of a person’s conduct strictly forming part of his private life, or even a complete and accurate picture of his private identity’ (Digital Rights Ireland Ltd v Minister for Communication , Opinion of AG Villalón: para 74).
Effective metrics for balancing the effectiveness and utility of new technological means of investigation and surveillance against the intrusiveness of the privacy invasion are in their infancy. The concept of privacy, therefore, is in urgent need of clarification in order to ensure that the laws designed to protect it are applied effectively and enabled to keep pace with technological innovation (Royal Academy of Engineering 2007; Metcalfe 2011: 99). The challenge facing scholars is to clarify the concept of privacy in this context. An initial task is to develop a list of indicators of intrusiveness.9 Several attempts have been made to gauge the relative intrusiveness (p. 721) of various potential intrusions on privacy by conducting surveys of public attitudes.10 Another approach would be to draw on the criminal law on battery. This draws on ‘background social norms’ to delineate the boundaries of personal space and determine the extent of any intrusion (Florida v Jardines 2013: 7). The offence of battery does not include ‘everyday touching’ or ‘generally acceptable standards of conduct’, for example bumping into someone on a crowded tube train or tapping someone on the shoulder to inform them that they have dropped something (Collins v Wilcock 1984). In a new world of digital communication in which so much of our personal lives are now conducted on digital formats (Roux and Falgoust 2013), these background norms are still in embryonic form.
There is an ‘obvious link’ between the right to privacy and the law on data protection and this relationship as well as the potential for safeguarding privacy in data protection law has been examined in detail elsewhere (Stoeva 2014).
Intelligence and law enforcement officials were largely untouched by early data protection regulation on account of broad exemptions for activities in pursuit of crime prevention and security in the Data Protection Directive 1995 (Bignami 2007) and the limited scope of the Council Framework Decision of 2008/977/JHA on the protection of personal data processed in the framework of police and judicial cooperation in criminal matters, which was restricted to the processing of personal data between member states.
The General Data Protection Regulation 2016/679 entered into force on 24 May 2016 and shall apply from 25 May 2018. Like its predecessor (the Data Protecting Direction 1995) the European Data Protection Regulation does not apply to law enforcement which is dealt with separately under the new data protection framework by Directive 2016/680. The Directive repeals Council Framework Decision 2008/977/JHA and EU Member States have until 6 May 2018 to transpose it into their national law. The new Directive does apply to national processing of data for law enforcement purposes as well as cross-border processing but scholars question ‘the police and justice sector being handled differently and separately from other sectors’ (Cannataci 2013). Debates about the extent to which the General Data Protection Regulation succeeds in grappling with the ‘computational turn of the 21st century’ and the ‘large-scale, complex, and multi-purpose forms of matching and mining zettabytes of data’ implied is of relevance to an increasingly automated criminal justice system (Koops 2013: 215).
There are clear parallels to be drawn between concerns expressed about police and court deference to science and technology in the criminal justice system and those exercising the drafters of Article 15(1) of the EC Data Protection Directive 1995 on the protection of individuals with regard to the processing of personal data. Article 15 granted the right to every person not to be subject to a decision which produces legal effects concerning them or significantly affects them and which is based solely on automated processing of data. It was the first attempt by data protection law to grapple directly with automated profiling.11 According to Article 12(a) (p. 722) of the Data Protection Directive 1995, each data subject had the right to obtain from the controller ‘knowledge of the logic involved in any automatic processing of data concerning him at least in the case of the automated decisions referred to in Article 15(1)’. That is
a decision which produces legal effects concerning him or significantly affects him and which is based solely on automated processing of data intended to evaluate certain personal aspects relating to him, such as his performance at work, creditworthiness, reliability, conduct, etc.
(articles 12(a) and 15(1) Data Protection Directive 1995).
The drafters of Article 15 of the 1995 Directive expressed concern that the ‘increasing automation of decision-making processes engenders automatic acceptance of the validity of the decisions reached and a concomitant reduction in the investigatory and decisional responsibilities of humans’ (Bygrave 2001: 18). A similar concern in relation to ‘push button forensics’ was discussed above and the same discomfort can be detected in judicial reluctance to convict defendants on the basis of statistical evidence alone, a prospect described as abhorrent by several evidence scholars on account of it being ‘the wrong kind of evidence’ (Duff 1998: 156). There is further resonance between the concerns underlying Article 15 and those motivating the protections enshrined in criminal procedure, which can be encapsulated in the concern to preserve human dignity with the principle of humane (or even human!) treatment. Article 15 was ‘designed to protect the interest of the data subject in participating in the making of decisions which are of importance to them’ (Commission of the European Communities 1990: 29). The rules of criminal procedure in the traditional criminal justice model share this objective. Recent decisions on hearsay in the European Court of Justice have reinvigorated debates on the meaning and importance of the ‘right to confrontation’ (O’Brian Jr 2011; Redmayne 2012). How well does this right—the right to confront one’s accuser—sit with the prospect of the ‘automated accuser’? (Wigan and Clarke 2013: 52).
The General Data Protection Regulation 2016/679 replicates and enhances the provisions regarding decisions based solely on automated processing (Articles 13, 21 and 22).
This element of data protection legislation has been described as particularly suited to enhance outcome/decision transparency, however it has been criticized for being confined in its application to decisions that are fully automated (Koops 2013: 211). The new obligation on data controllers to provide information as to the existence of processing for an automated decision and about the significance and envisaged consequences has been described as providing a potentially ‘revolutionary’ step forwards in levelling the playing field in profiling practices (Koops 2013: 200). However, the Regulation has been criticized for confining its focus to ex ante process transparency and for failing to extend its reach to ex post outcome/event/decision transparency (Koops 2013: 200). Greater decision transparency is advocated for ‘in the form of understanding which data and which weight factors (p. 723) were responsible for the outcome’ on the basis that it is this, as opposed to the subject’s awareness of their participation in the provision of data, that would enable a decision based upon its analysis to be effectively challenged or revised (Koops 2013: 213).
Article 11 of the 2016 Directive (data protection processing by law enforcement) obliges member states to prohibit decisions based solely on automated processing including profiling, which produce and adverse or other ‘significant affect’ in the absence of lawful authorisation containing ‘appropriate safeguards’ (including ‘at least the right to obtain human intervention on the part of the controller’) and prohibits profiling that results in discrimination on the basis of certain special categories of personal data (article 11 of the 2016 Directive). It is not clear what additional appropriate safeguards will be nor whether there are any further obligatory safeguards. Should these safeguards only apply where the decision is based solely on automated processes? What about where the decision is based on semi-automated processes and human judgement? It is not clear what will be deemed to be a ‘significant affect’ under the Directive. Where the consquence of the processing is an increased risk of being searched by the police, will this amount to a ‘significant affect’? Might the exercise of human judgment in relation to whether to act on the data and conduct a search mean that the decision is not based soley on automated processing? The Directive make no reference to any obligation to disclose the logic of automated processing. Shouldn’t citizens have a right to be informed of the logic behind decisions to flag their vehicles as suspicious? (Koops 2013)
Recent changes to the Criminal Procedure Rules on expert evidence in England and Wales (along the lines recommended by the Law Commission) seek to ensure that wherever expert opinion is admitted at trial, the basis for that opinion is laid bare in a manner comprehensible to lay persons. Where alternative disposal methods are invoked, decision transparency is obfuscated and as discussed above, particularly where the inferences drawn are based on complex algorithms, even those exercising the decision may be ignorant of the details of the inferential process used. In the absence of algorithmic transparency miscarriages of justice will be more difficult to detect.
7.3 Incorporating Criminal Justice Values within the Rapidly Evolving and Heavily Contested Field of Data Protection Law
There is widespread misunderstanding and ambiguity surrounding the extent to which data protection regulations apply to law enforcement (O’Floinn 2013) but data protection bodies are increasingly drawing on the ‘due process’ values developed over many years in the theory and practice of the criminal law. The UK (p. 724) Information Commissioner recently acknowledged that the principle of fair and lawful processing of data is engaged when a public statement by the police undermines the presumption of innocence. In a statement issued by the ICO, a spokesperson explained:
The ICO spoke to Staffordshire Police following its #DrinkDriversNamedonTwitter campaign. Our concern was that naming people who have only been charged alongside the label ‘drink driver’ strongly implies a presumption of guilt for the offence, which we felt wouldn’t fit with the Data Protection Act’s fair and lawful processing principle.
Such a development in the application of data protection principles could go some way towards addressing concerns expressed by Campbell about the procedural impediments to expanding the presumption of innocence to state practices outside of the traditional model such as inclusion on ‘watch lists’, and publication of civil preventative orders (Campbell 2013). Although in the case dealt with by the Information Commissioner, the persons named had been charged and were awaiting trial, the fact that the spirit of the presumption of innocence could be incorporated within the fair and lawful processing principle is a promising development.12
Historically, the parameters of government control over the citizen have been most clearly articulated in the rules of criminal procedure and evidence. This is largely on account of the criminal law having traditionally provided the state with its most coercive and intrusive powers and mechanisms with which to carry out surveillance, control and punishment (Bowling, Marks, and Murphy 2008). Recent technological innovations have exacerbated tensions in the traditional model of criminal justice to the point that its internal and external boundaries are on the brink of collapse. New scientific techniques, data collection devices, and mathematical analytical procedures are having a profound effect on the administration of criminal justice. They are blurring the boundary between the innocent person who should be able to expect freedom from state intrusion and coercion and the ‘reasonably suspected’ person for whom some rights may justifiably be curtailed. These same technologies are also blurring the boundary between the accused and the convicted. The established process that distinguishes the collection of evidence, the testing of its veracity and probative value and the adjudication of guilt are being automated and temporally and procedurally compressed. At the same time, the start and finish of the criminal justice process are now (p. 725) indefinite and indistinct as a result of the introduction of mass surveillance and the erosion of ‘double jeopardy’ protections caused by scientific advances that make it possible to revisit conclusions reached in the distant past. In our opinion, this drift is endangering the privacy and freedom of all citizens and most perniciously that of the ‘usual suspects’. There may be scope for protection in expanding the application of criminal justice safeguards to everyone. However, it may be that the speed and depth of the technological changes occurring in this area are so great that the traditional criminal justice model—with such precepts as the presumption of innocence, the separation between the collection of evidence by the police and the testing of that evidence by a court of law, and so on—is no longer fit for the purpose of explaining criminal justice practice or constraining state power. Our deepest concern is the emergence of a potentially unfettered move towards a technologically driven process of ‘automatic criminal justice’. It may be that a stronger right to privacy and enhanced data protection rights could prove a more solid foundation for building a model that will protect fundamental human rights and civil liberties in the long term. We think that in an increasingly digital world, further exposition of traditional criminal justice values is required along with a detailed examination of how these values can be combined with data protection developments to provide proper safeguards in the new technologically driven world of criminal justice.
Alldridge P, ‘Do C&IT Facilitate the Wrong Things?’ (2000) 14 International Rev of L, Computers & Technology 143Find this resource:
Alldridge P and C Brants, Personal Autonomy, the Private Sphere and Criminal Law: A Comparative Study (Hart 2001)Find this resource:
Ashworth A and M Redmayne, The Criminal Process (4th edn, OUP 2010)Find this resource:
Baines J, ‘Staffs Police to Drop Controversial Naming “Drink Drivers” Twitter Campaign’ (Information Rights and Wrongs, 2014) <http://informationrightsandwrongs.com/2014/01/23/staffs-police-to-drop-controversial-naming-drink-drivers-twitter-campaign/> accessed 25 October 2015Find this resource:
Bauman Z and others, ‘After Snowden: Rethinking the Impact of Surveillance’ (2014) 8 International Political Sociology 121Find this resource:
Bignami F, ‘Privacy and Law Enforcement in the European Union: The Data Retention Directive’ (2007) 8 Chicago Journal of International Law 233Find this resource:
Borts P, ‘Privacy, Autonomy and Criminal Justice Rights’ in Alldridge and Brants (2001)Find this resource:
Bosco F and others, ‘Profiling Technologies and Fundamental Rights and Values: Regulatory Challenges and Perspectives from European Data Protection Authorities’ in Serge Gutwirth, Ronald Leenes, and Paul de Hert, Reforming European Data Protection Law (Springer 2015)Find this resource:
Bowling B, A Marks, and C Murphy, ‘Crime Control Technologies: Towards an Analytical Framework and Research Agenda’ in Roger Brownsword and Karen Yeung (eds), Regulating Technologies: Legal Futures, Regulatory Frames and Technological Fixes (Hart 2008)Find this resource:
Brodeur J, ‘High Policing and Low Policing: Remarks about the Policing of Political Activities’ (1983) 30 Social Problems 507Find this resource:
Bygrave L, ‘Automated Profiling: Minding the Machine: Article 15 of the EC Data Protection Directive and Automated Profiling’ (2001) 17 Computer Law and Security Review 17Find this resource:
(p. 727) Campbell L, ‘A Rights-Based Analysis of DNA Retention: “Non-Conviction” Databases and the Liberal State’ (2010) 12 Criminal L Rev 889Find this resource:
Campbell L, ‘Criminal Labels, the European Convention on Human Rights and the Presumption of Innocence’ (2013) 76 Modern Law Review 681Find this resource:
Cannataci J, ‘Defying the Logic, Forgetting the Facts: The New European Proposal for Data Protection in the Police Sector’ (2013) 4(2) European Journal of Law and TechnologyFind this resource:
Clarke R, ‘Profiling: A Hidden Challenge to the Regulation of Data Surveillance’ (1993) 4 Journal of Law and Information Science 403Find this resource:
Collins v Wilcock  1 WLR 1172
Commission of the European Communities, ’Proposal for a Council Directive concerning the protection of individuals in relation to the processing of personal data’ COM(90) 314 final ~ SYN 287, 13 September 1990 (1990)Find this resource:
Council Directive 95/46/EC of 24 October 1995 on the protection of individuals with regard to the processing of personal data and on the free movement of such data  OJ L281/31 (Data Protection Directive)Find this resource:
Damaška MR, Evidence Law Adrift (Yale UP 1997)Find this resource:
Deleuze G and F Guattari, A Thousand Plateaus: Capitalism and Schizophrenia (Athlone Press 1988)Find this resource:
Dennis I, ‘Rethinking Double Jeopardy: Justice and Finality in Criminal Process’  Crim LR 933Find this resource:
Department of Constitutional Affairs and Attorney General’s Office, ‘Delivering Simple, Speedy, Summary Justice’ (2006)Find this resource:
Digital Rights Ireland Ltd v Minister for Communication  Cases C-293/12 and C-594/12 (Opinion of AG Villalón)
Dewan S, ‘Judges Replacing Conjecture with Formula for Bail’ (New York Times, 26 June 2015) <www.nytimes.com/2015/06/27/us/turning-the-granting-of-bail-into-a-science.html> accessed 25 October 2015Find this resource:
Dhami M and P Ayton, ‘Bailing and Jailing the Fast and Frugal Way’ (2001) 14 Journal of Behavioral Decision Making 141Find this resource:
Duff R, ‘Dangerousness and Citizenship’ in Andrew Von Hirsch, Andrew Ashworth and Martin Wasik (eds), Fundamentals of Sentencing Theory: Essays in Honour of Andrew von Hirsch (Clarendon Press 1998)Find this resource:
Duff R and others, The Trial on Trial: Towards a Normative Theory of the Criminal Trial, Volume 3 (Hart 2007)Find this resource:
Eckes C, EU Counter-Terrorist Policies and Fundamental Rights: The Case of Individual Sanctions (OUP 2009)Find this resource:
Ericson R and K Haggerty, Policing the Risk Society (OUP 1997)Find this resource:
Feeley M, ‘Origins of Actuarial Justice’ in Sarah Armstrong and Lesley McAra (eds), Perspectives on Punishment: The Contours of Control (OUP 2006)Find this resource:
Feeley M and J Simon, ‘The New Penology: Notes on the Emerging Strategy of Corrections and Its Implications’ (1992) 30 Criminology 449Find this resource:
Feeley M and J Simon, ‘Actuarial Justice: The Emerging New Criminal Law’ in David Nelken (ed), The Futures of Criminology (SAGE 1994)Find this resource:
Fenton N, ‘Effective Bayesian Modelling with Knowledge before Data’ (2013) ERC Advanced Grant 2013 Research proposal [Part B1]Find this resource:
Florida v Jardines, 569 US (2013)
(p. 728) Forensic Access, ‘Streamlined Forensic Reporting (SFR)—Issues and Problems’ (2014) <www.forensic-access.co.uk/streamlined-forensic-reporting-sfr-problems-issues.asp> accessed 19 May 2014
Foucault M, Discipline and Punish: The Birth of the Prison (Pantheon 1977)Find this resource:
Garland D, ‘ “Governmentality” and the Problem of Crime: Foucault, Criminology, Sociology’ (1997) 1 Theoretical Criminology 173Find this resource:
Garland D, The Culture of Control: Crime and Social Order in Contemporary Society (OUP 2001)Find this resource:
Haggerty K, ‘Technology and Crime Policy’ (2004) 8 Theoretical Criminology 491Find this resource:
Haggerty K and R Ericson, ‘The Surveillant Assemblage’ (2000) 51 British Journal of Sociology 605Find this resource:
Hilbert M, ‘How Much Information Is There in the “Information Society”?’ (2012) 9 Significance 8Find this resource:
Hildebrandt M, ‘Ambient Intelligence, Criminal Liability and Democracy’ (2008) 2 Criminal Law and Philosophy 163Find this resource:
Hogan-Howe B, ‘2020 Vision: Public Safety in a Global City’ (Speech at Royal Society of Arts, 12 March 2015)Find this resource:
Hudson B, Justice in the Risk Society: Challenging and Re-affirming ‘Justice’ in Late Modernity (SAGE 2003)Find this resource:
Iszatt A, ‘Fingerprints: The Path to the Soul’ (Police Oracle, 2 May 2014) <www.policeoracle.com/news/Investigation/2014/May/02/Fingerprints-The-path-to-the-soul_81363.html/technology> accessed 25 October 2015Find this resource:
Jackson J, ‘ “Police and Prosecutors after PACE”: The Road from Case Construction to Case Disposal’ in Ed Cape and Richard Young (eds), Regulating Policing: The Police and Criminal Evidence Act 1984 Past, Present and Future (Hart 2008)Find this resource:
James J and P Gladyshev, ‘Challenges with Automation in Digital Forensic Investigations’ (2013) <http://arxiv.org/abs/1303.4498> accessed 25 October 2015
Joh E, ‘Policing by Numbers: Big Data and the Fourth Amendment’ (2014) 89 Washington L Reform 35Find this resource:
Jones C, ‘Auditing Criminal Justice’ (1993) 33 British Journal of Criminology 187Find this resource:
Kerr I and J McGill, ‘Emanations, Snoop Dogs and Reasonable Expectations of Privacy’ (2007) 52 Criminal Law Quarterly 392Find this resource:
Kinsella C and J McGarry, ‘Computer says No: Technology and Accountability in Policing Traffic Stops’ (2011) 55 Crime, Law and Social Change 167Find this resource:
Koops B-J, ‘On Decision Transparency, or How to Enhance Data Protection after the Computational Turn’ in Mireille Hildebrandt and Katja de Vrie (eds), Privacy, Due Process and the Computational Turn: The Philosophy of Law Meets the Philosophy of Technology (Routledge 2013)Find this resource:
Law Commission, Double Jeopardy, Consultation Paper No 156 (1999) 37Find this resource:
Law Commission, Expert Evidence in Criminal Proceedings in England and Wales (Law Com No 325, 2011)Find this resource:
Leopold G, ‘Can Big Data Help Dispense Justice?’ (Datanami, 12 December 2014) <www.datanami.com/2014/12/12/can-big-data-help-dispense-justice/> accessed 25 October 2015Find this resource:
Lyon D, Surveillance as Social Sorting: Privacy, Risk, and Digital Discrimination (Psychology Press 2003)Find this resource:
McCartney C, ‘Universal DNA Database Would Make Us All Suspects’ New Scientist (19 September 2007)Find this resource:
(p. 729) McGarry J, ‘Named, Shamed, and Defamed by the Police’ (2011) 5 Policing 219Find this resource:
Macnish K, ‘Unblinking Eyes: The Ethics of Automating Surveillance’ (2012) 14 Ethics and Information Technology 151Find this resource:
Maguire M, ‘Policing by Risks and Targets: Some Dimensions and Implications of Intelligence‐led Crime Control’ (2000) 9 Policing and Society: An International Journal 315Find this resource:
Marks A, ‘Expert Evidence of Drug Traces: Relevance, Reliability and the Right to Silence’ (2013) 10 Criminal L Rev 810Find this resource:
Marx G, ‘Technology and Social Control: The Search for the Illusive Silver Bullet’ (2001) International Encyclopedia of the Social and Behavioral SciencesFind this resource:
Marx G, ‘Surveillance and Society’ in Ritzer G (ed), Encyclopedia of Social Theory (SAGE 2005)Find this resource:
Marx G, ‘The Engineering of Social Control: Policing and Technology’ (2007) 1 Policing 46Find this resource:
Metcalfe E, Freedom from Suspicion: Surveillance Reform for a Digital Age: A Justice Report (Justice 2011)Find this resource:
Mitsilegas V, ‘The Value of Privacy in an Era of Security: Embedding Constitutional Limits on Preemptive Surveillance’ (2014) 8 International Political Sociology 104Find this resource:
Motor Insurers’ Bureau, ‘Welcome to the Motor Insurers Bureau’ (2010) <www.mib.org.uk/Home/en/default.htm> accessed 8 July 2015
Murphy E, ‘The New Forensics: Criminal justice, False Certainty, and the Second Generation of Scientific Evidence’ (2007) 95 California L Rev 721Find this resource:
O’Brian W, Jr ‘Confrontation: The Defiance of the English Courts’ (2011) 15 International Journal of Evidence and Proof 93Find this resource:
O’Floinn M, ‘It Wasn’t All White Light before Prism: Law Enforcement Practices in Gathering Data Abroad, and Proposals for Further Transnational Access at the Council of Europe’ (2013) 29 Computer L and Security Rev 610Find this resource:
O’Malley P, ‘Volatile and Contradictory Punishment’ (1999) 3 Theoretical Criminology 175Find this resource:
Orwell G, 1984 (Secker & Warburg 1949)Find this resource:
Osborne D and T Gaebler, Reinventing Government: How the Entrepreneurial Spirit is Transforming Government (Addison-Wesley 1992)Find this resource:
Packer H, The Limits of the Criminal Sanction (Stanford UP 1968)Find this resource:
Palmer A, ‘When It Comes to Sentencing, a Computer Might Make a Fairer Judge Than a Judge’ (The Telegraph, 21 January 2012) <www.telegraph.co.uk/news/uknews/law-and-order/9029461/When-it-comes-to-sentencing-a-computer-might-make-a-fairer-judge-than-a-judge.html> accessed 2 November 2015Find this resource:
Ramsey M, ‘A Return Flight for Due Process? An Argument for Judicial Oversight of the No-Fly List’ (2014) <http://ssrn.com/abstract=2414659> accessed 4 November 2015
Redmayne M, ‘Confronting Confrontation’ in Paul Roberts and Jill Hunter (eds), Criminal Evidence and Human Rights: Reimagining Common Law Procedural Traditions (Hart 2012)Find this resource:
R v Dyment  2 SCR 417
Rhodes RA, ‘The Hollowing Out of the State: The Changing Nature of the Public Service in Britain’ (1994) 65 Political Quarterly 138Find this resource:
Roberts P, ‘Double Jeopardy Law Reform: A Criminal Justice commentary’ (2002a) 65 MLR 93Find this resource:
Roberts P, ‘Justice for All? Two Bad Arguments (and Several Good Suggestions) for Resisting Double Jeopardy Reform’ (2002b) 6 E&P 197Find this resource:
(p. 730) Roberts P and Zuckerman A, Criminal Evidence (2nd edn, OUP 2010)Find this resource:
Rose N, ‘Government and Control’ (2000) 40 British Journal of Criminology 321Find this resource:
Rose N and P Miller, ‘Political Power beyond the State: Problematics of Government’ (1992) 43 British Journal of Sociology 173Find this resource:
Roux B and M Falgoust, ‘Information Ethics in the Context of Smart Devices’ (2013) 15 Ethics and Information Technology 183Find this resource:
Royal Academy of Engineering, ‘Dilemmas of Privacy and Surveillance: Challenges of Technological Change’ (2007)Find this resource:
S and Marper v United Kingdom (2009) 48 EHRR 50
Select Committee on the Constitution, Surveillance: Citizens and the State (HL 2008–2009, 18-I)Find this resource:
Sherman L, ‘The Rise of Evidence-based Policing: Targeting, Testing, and Tracking’ (2013) 42 Crime and Justice 377Find this resource:
Slobogin C, Privacy at Risk: The New Government Surveillance and the Fourth Amendment (University of Chicago Press 2007)Find this resource:
Solove D, ‘Data Mining and the Security–Liberty Debate’ (2008) 75 University of Chicago Law Review 343Find this resource:
Stoeva E, ‘The Data Retention Directive and the right to privacy’ (2014) 15 ERA Forum 575–592Find this resource:
Strange S, The Retreat of the State: The Diffusion of Power in the World Economy (CUP 1996)Find this resource:
Stuckenberg, C-F, ‘Who Is Presumed Innocent of What by Whom’ (2014) 8 Crim Law and Philos 301–316Find this resource:
Sullivan G and B Hayes, Blacklisted: Targeted Sanctions, Preemptive Security and Fundamental Rights (European Center for Constitutional and Human Rights 2011)Find this resource:
Van Dijck J, ‘Datafication, Dataism and Dataveillance: Big Data between Scientific Paradigm and Ideology’ (2014) 12 Surveillance and Society 197Find this resource:
Wey T, DT Blumstein, W Shen, and F Jordán, ‘Social Network Analysis of Animal Behaviour: A Promising Tool for the Study of Sociality’ (2008) 75 Animal Behaviour 333Find this resource:
Wigan M and R Clarke, ‘Big Data’s Big Unintended Consequences’ (2013) 46 Computer 46Find this resource:
Yassin Abdullah Kadi and Al Barakaat International Foundation v Council and Commission  ECR I-6351
Young R, ‘Street Policing after PACE: The Drift to Summary Justice’ in Ed Cape and Richard Young R (eds), Regulating Policing: The Police and Criminal Evidence Act 1984 Past, Present and Future (Hart 2008)Find this resource:
Zedner L, ‘Security, the State, and the Citizen: The Changing Architecture of Crime Control’ (2010) 13 New Criminal Law Review 379Find this resource:
(*) The work of Amber Marks was supported in part by the European Research Council (ERC) through project, ERC-2013-AdG339182-BAYES_KNOWLEDGE.
(2.) For the UK, see: Law Commission Consultation Paper No 190, ‘The admissibility of expert evidence in criminal proceedings in England and Wales: a new approach to the determination of evidentiary reliability (2009); Law Commission Report No 325 ‘Expert evidence in criminal proceedings in England and Wales’ (2011); House of Commons Science and Technology Committee, Forensic Evidence on Trial, Seventh Report of Session 2004–05 p 76; The Fingerprint Inquiry (Scotland, 2011). For Canada, see Goudge ST. Inquiry into paediatric forensic pathology in Ontario. Toronto (ON): Ontario Ministry of the Attorney General, 2008.
(4.) Criminal Justice Act 2003, Part 10.
(5.) We have borrowed this terminology from Mirelle Hildebrandt’s discussion of how to preserve the achievements of constitutional democracy in a world of ambient intelligence (2008: 167).
(6.) For a detailed unpacking of the presumption of innocence including an exploration of the basic notion of ‘innocence’ as well as the historical role of the presumption in shielding the accused from the hardships of the criminal process, see Stuckenberg (2014).
(11.) Article 15 grants the right to every person not to be subject to a decision which produces legal effects concerning them or significantly affects them and which is based solely on automated processing of data intended to evaluate certain personal aspects relating to them.
(12.) See also Civil Rights Principles for the Era of Big Data, ACLU, 2014 available at <http://www.civilrights.org/press/2014/civil-rights-principles-big-data.html> (accessed 2 July 2015).