key: cord-0867810-xo9b5dl5 authors: Loo, Jane; Findlay, Mark title: Digitised Justice: The New Two Tiers? date: 2022-02-27 journal: Crim Law Forum DOI: 10.1007/s10609-022-09431-x sha: d0a157ad95364cd3b7a130669d48d79a63f12438 doc_id: 867810 cord_uid: xo9b5dl5 Prevailing conditions of access to justice and due process in the Singapore courts are criticised through McBarnet’s two-tier lens and Carlen’s dramaturgical understandings of criminal court realities. More than an interest in the structural separation of the Singapore judiciary, the paper interrogates the dualism between the imagined workings of justice and the daily operational experience for users of the Singapore courts. The scene is set to understand ideologies of triviality and irrelevance and their impact on justice service delivery in subordinate courts where legal representation and offender participation are the exception. To speculate on novel influences of triviality and irrelevance through machine-learned automation, an audit of Singapore’s present-day court technologies and its increasingly digitised court processes and format is detailed in the second part. The administrative benefits of digitisation notwithstanding, the paper reasons that digitalisation motivated by convenience, cost-cutting and emergency exigencies presents additional dangers to justice access and due process delivery above those already at play. These further challenges are deciphered through considerations of how justice service delivery is depersonalised and routinised in disruptive digital models. Digitized justice suggests a new ‘two tiers’ duality between physical and virtual frames of service delivery and contestation. The ‘on-line’ screen shifts the theatre from the courtroom to the ‘zoom room’. This exploration of two tiers of justice and theatre of the courtroom re-imagined through digitisation offers the opportunity to appreciate and activate automated decision processes and data management as part of the solution, rather than conceding their exacerbation of the ‘injustice’ posed by this two tiers ideology and courtroom drama exclusionism. In her report on post-pandemic justice possibilities, Liviana Zorzi of the UNDP observes that the conditions of pandemic control have increased the need for justice intervention in many crucial areas of social engagement. Nevertheless, the demand for justice has not decreased; on the contrary, the number of people's justice problems is increasing rapidly as they lose jobs and run into difficulties paying bills and debts; a steep increase of business disputes related to bankruptcy and insolvency is expected, while domestic violence cases are already on the rise in several countries. 1 Zorzi suggests that the pandemic is worsening an already dysfunctional gap in access to justice resolutions. With courts turning more now to digital tools and almost half the worlds' population without internet connection, there is a pressing need to place universal access to justice at the forefront of advances in digitised modes since the unequal/lack of digitised access will threaten public confidence in the promise of due process and justice for all. Recognition of due process is also exemplified by the spirit of Sustainable Development Goal 16: ''peace, justice, and sustainable institutions''. SDG 16 identifies the promotion of access to justice as a high developmental priority 2 and as the roadmap to building back better 3 post-pandemic. Access to justice is more than just making available to marginalized individuals, judicial and administrative institutions, and processes. 4 Access to justice is only realized when a greater number of individuals are offered convenient justice pathways in the context of due process service delivery and institutional accountability. Such conditions for justice are particularly crucial for the indigent, the disadvantaged, and those scarred from their engagement with justice administration. In recent years there has been much debate about whether the introduction of AI-assisted technology into legal service delivery aids or impairs access to justice. 5 These developments have on occasions been criticized as being motivated by administrative efficiencies and cost savings that worsen the quality of justice delivery. The exigencies of the pandemic have meant that many mundane court processes have gone online. 6 Even prior to COVID, many countries have been experimenting with digitised justice from less controversial applications such as automated case management to more problematic explorations of predictive sentencing. 7 This process has preceded, and no doubt will post date the pandemic. The paper advances a critique of digitised justice measured against access to justice, 8 using the Singapore court process as a case study. The paper describes aspects particularly of the subordinate courts and trial process that might at present challenge access (both structural and process impediments). Moving from these realities, the paper suggests how digitisation motivated primarily by administrative priorities, raises dangers to access above those currently present. To reposition the reality of digitising in justice service delivery, the analysis concludes by proposing how access to justice can be lifted to a central priority in digitisation. Two tiers of justice and dramaturgical understandings of trial justice: The theoretical foundation for the analysis rests on an elaboration of MacBarnet's two tiers of justice in the contemporary context of active and impending digitisation. It takes Ôtwo tiers' from the duality of court hierarchies and moves it to the virtual interface of digitized justice today. In seeking to translate Ôtwo tiers' into contemporary Singapore settings, certain incompatibilities require thought. First, McBarnet was writing in an age when judges manually recorded hearings and wrote out their deliberations long-hand, with little to no incursion of technology in court processing. Additionally, the Scottish court structure of Magistrate's justice in MacBarnet's research at the time does not translate and apply wholesale into the court divisions in Singapore. Even so, we argue that the two tiers dichotomy should be pursued as a worthwhile theoretical base for the critique of digitisation. Two tiers remains instructive in identifying the initial ideological inhibitions of triviality and legal irrelevance that could further jeopardize claimants' access to justice in the digitised model. Digitalisation additionally contributes to the routinising and depersonalising of processes and institutions. 9 These two additional theoretical param- 8 The paper acknowledges that it would be inaccurate and unfair to suggest that only digitising is a challenge to equitable access. 9 Briefly, depersonalisation is the transition towards machine learning, predictive technologies, automated interaction and online deliberation from previous humancentric justice and more personal justice/court interactions. Justice is routinised and made mechanical because of utilitarian cost/benefit approaches taken towards the sustainability and suitability of justice service delivery. When this Ôeconomic analysis' lens is prioritized it excludes the subject from participation in their own proceedings. eters draw on Carlen's theatre of the courtroom 10 to suggest another way of approaching two tiers and due process; the distinction between popular culture imaginings of due process courtroom justice, and what happens in practice. Digital justice delivery represents a new dramaturgical medium where participation is mediated through computerised technology and separated by the digital screen. The exploration of two tiers and dramaturgical understandings of trial justice in the Singapore context is meant to challenge the uniform availability of due process and therefrom, the nature of access to justice in this digital age. If, as the analysis to follow suggests, digitisation can reframe due process and expand the divide between justice seen to be done and juridical practice achieved through technology, then it is fitting to suggest that our digitising priorities ought to be re-examined. In 1981 Doreen McBarnet published her article ÔMagistrates' Courts and the Ideology of Justice', 11 in which she posed the ideological problems reflected in the institutional bases of lower courts justice. If the lower courts seem to present a different world from the image we carry in our heads from the higher courts, then it is hardly surprising; in law that is exactly what they are. The law has created two tiers of justice, one which is geared in its ideology and generality at least to the structures of legality, and one which quite simply and explicitly is not. 12 From this observed duality, she progressed to analyse how differential structures of trial justice are legitimized. First, both the offences and the penalties involved in magistrates' justice are too trivial for the structures of due process; second, the issues and processes are such that the niceties of law and lawyers are irrelevant. 13 There is an inherent paradox in the very idea of prosecuting trivial offences. They are too trivial to interest the public but not too trivial for the state to prosecute in the name of the public; too trivial to merit due process of law but not too trivial for the intervention of the law. The ideology of triviality focuses on the offences and penalties, not on the question of prosecution itself. […] it is exactly in the area of minor offences that the operation of the law, in terms of democratic justice, becomes most suspect. […] the lower courts remain something to be laughed at or yawned over for the pettiness of their crimes, not watched with care for the marginality of their legality. 14 As with the second justification for reducing the structures of due process in the lower courts, legal relevance, McBarnet critiques: The image of the lower courts as not needing lawyers, which justifies not providing lawyers, is itself a product of their absence. The defendant is thus caught up in the vicious circle that lies behind the image of ''simple facts'' and ''straightforward cases''. […] the profession too is imbued with the dominant images of the lower courts as neither serious enough nor legally relevant enough to need lawyers. 15 In her conclusion, McBarnet reflects on how these structural differences in the delivery of legal services are not simply ideological accomplishments but provide ideological functions. Legal policy has established two tiers of justice. One, the higher courts, is for public consumption, the arena where the ideology of justice is put on display. The other, the lower courts, deliberately structured in defiance of the ideology of justice, is concerned less with subtle ideological messages than with direct control. The latter is closeted from the public eye by the ideology of triviality, so the higher courts alone feed into the public image of what the law does and how it operates. But the higher courts deal with only 2% of the cases that pass through the courts. Almost all criminal law is acted out in the lower courts without traditional due process. […] The 98% becomes the exception to the rule of ''real law'' and the working of the law comes to be typified not by its routine nature, but by its atypical, indeed exceptional high court form. Between them the ideologies of triviality and legal irrelevance accomplish the remarkable feats of defining 98% of court cases not only as exceptions to the rule of due process, but also as of no public interest whatsoever. 16 With the onset of digitised justice, the spectre of two tiers takes on a different structural form but supported by the same ideological frame. 17 through the court ecosystems prior to digitisation will predetermine the manner in which automated systems will then be designed and realized. Thus, if certain contestations and contestants are treated as trivial, and their interests regarded as less relevant for law, then the natural consequence will be precisely that; digitalisation and its properties of depersonalisation and routinisation will worsen conditions of irrelevance and triviality in justice service delivery. From this design environment, any resultant digitized justice has the potential to exacerbate access and due process impediments that already exist, providing even less productive participation. Digitising various aspects of procedural and substantive justice have already been underway for decades and the necessities of social distancing as part of the COVID 19 pandemic control policies have pushed the digitising of justice on apace. 18 Many of these developments are being achieved without sufficient reflection on the potential exacerbation of two tiers of justice and how justice is affected in practice. Even in jurisdictions such as Singapore where the tiers are not so distinct from lower to higher courts, there remains a dualism between the imagined workings of justice and the daily operational experience. Certain stakeholders are at risk of further depersonalised and routinised justice service delivery justified by existing perceived triviality and irrelevance. Moving on to The Staging of Magistrates Justice, 19 Pat Carlen described what she researched in the lower courts as; Ôproceedings with a surrealism which atrophies defendants' ability to participate in them.' 20 She categorised what the observer witnessed of lower court justice in terms of space (ramped), time (or lack of it), and presentations (inter-professionality and exclusion of lay participants). Her thesis is that through the structural format and processual distraction of courtroom regimes, the Ôtheatre of the absurd' plays out for the benefit of anyone but the accused person. 21 21 Theatre of the absurd here relates to its four central characteristics: anti-character, anti-language, anti-plot and anti-drama. It is playing out the absurdity of human existence in a meaningless universe. Though structurally opposed, the theatre of the absurd and the court of law have several phenomenological features in common. Their central divergence inheres in their opposed structural functions. Thus, whereas dramatists of the absurd intentionally and overtly utilise the plausible and the mundane to construct the overtly senseless and absurd, the mandarins of justice intentionally and covertly utilise the plausible and mundane to construct the covertly senseless and absurd. In magistrates' courts, as in the theatre of the absurd, mundane and conventional ways of organising and communicating the operative meanings of social occasions are simultaneously exploited and denied. Yet their outcomes are situationally authenticated and the intermeshed structures of surrealism and psychic coercion are difficult to locate. This is because police and judicial personnel systematically present their coercive devices as being nothing more than the traditional, conventional and commonsensical ways of organising and synchronising judicial proceedings. 22 In the reality of the defendant's experience, for which this theatre should primarily be played out, the language is impenetrable, the characters foreign, the drama unapproachable, and the plot is a fabrication created by professionals and their questioning. Shift this from the physical premises of austere courtrooms to the perspectives of the computer screen and the absurdity for the lay participant becomes even more extreme. Through digitising justice, the spatial and temporal pre-determinants of courtroom theatre are made more mundane and exclusionist in that the players in the drama are depersonalised, and the space, time, and presentation components are all routinised through one more window of separation: the computer screen or the video monitor. In this context, the defendant is further at risk of exclusion with no temporal or spatial anchors to govern their interpretative capacity. Additionally, their chance for participation and communication is also impacted if the controllers of the online facility present further barriers to interactive access. The critical description to follow exists in three parts. The first describes the current challenges to access and due process in the Singapore criminal justice process at present. In this section there is a regular reflection on McBarnet's ideological tools of triviality and irrelevance, and Carlen's theatre of the absurd. Where appropriate, brief projections on the impact of digitising relative to routinising and 22 Carlen (n 25).49. depersonalising are explained. The second section looks at the present state of digitised justice in Singapore as a precursor to critical reflections concerning these developments and further digitising. The final part examines the impact of digitisation and its construction of depersonalised and routinised justice; this will be judged against impediments to access to justice and due process priorities. Singapore has two tiers of courts, the State Courts and the Supreme Court. 23 The State Courts comprise of both the District Courts and Magistrate Courts which oversee criminal, as well, civil matters. The District Courts hear criminal cases where the maximum imprisonment term does not exceed 10 years, or which are punishable with a fine only. 24 The Magistrate Courts hear criminal cases where the maximum imprisonment term does not exceed 5 years, or which are punishable with a fine only. 25 The Supreme Court of Singapore consists of the High Court and the Court of Appeal. 26 The High Court has a General Division and an Appellate Division. The General Division of the High Court exercises original and appellate jurisdiction in civil and criminal cases. 27 It also has revisionary jurisdiction over the State Courts in criminal matters. The Appellate Division of the High Court has no criminal jurisdiction. 28 So that the utility of Ôtwo tiers' theorising and the theatre of the courtroom can transit into virtual justice frames it is necessary first to interrogate these theoretical underpinnings in their Ôpre-virtual' analysis. At the outset, it is observed that approximately 90% of the 23 ÔSingapore Judicial System' (Supreme Court Singapore) accessed 17 February 2021. 24 ÔJurisdiction of the Supreme Court' (Supreme Court Singapore) accessed 17 February 2021. 25 ÔJurisdiction of the Supreme Court' (n 30). 26 ÔSingapore Judicial System' (n 29). 27 ÔSingapore Judicial System' (n 29). 28 ÔSingapore Judicial System' (n 29). Singapore Judiciary's caseload and 99% of criminal cases in Singapore are heard in the State Courts. 29 As stated, the distinction between due process exercised in the Singapore State Courts and the High Court for criminal matters may not appear as obvious as was apparent in McBarnet's study. Low levels of legal representation are comparable across the court hierarchy in Singapore, and attitudes to the complexity and seriousness of criminal litigation may not vary so markedly from State to High Courts. This lesser distinctiveness could be a factor of clear discriminators in higher court trial practice such as juries being absent from the Singapore courts. Nevertheless, the specific manifestations of the ideology of triviality and of the lack of legal relevance that impact on due process at the State Courts level can still be gathered from the treatment of accused persons in the following areas and procedures: the practice of trial waivers and guilty pleas, the number of unrepresented accused persons at trial and the delay in access to counsel on arrest. Ultimately, the two-tier duality in Singapore may not be represented in the trappings of legality but rather, the extent 30 to which in the subordinate courts' triviality and irrelevance reduce the relationship between the accused and due process to its most minimal identifiers. 29 The Honourable Justice Belinda Ang Saw Ean Judge of the Supreme Court of Singapore, ÔSMU Forum -Expanding the Scope of Dispute Resolution and Access to Justice: The Use of Mediaation within the Courts'; (18 January 2018) accessed 24 February 2021. See also: Priyankar Bhunia, ÔState Courts of Singapore and A*STAR's Institute for Infocomm Research Work ing on Real-Time Speech Transcription System' (Open Gov Asia, 16 December 2017) accessed 24 February 2021. 30 A note first on the availability of empirical data underpinning our research in this section -statistics and percentages relevant to the (1) number of cases disposed of via guilty pleas in the State Courts (2) number of unrepresented accused persons in the State Courts (3) number of cases heard at the State Courts without access to legal aid and (4) conviction and acquittal rates in the State Courts have not been made publicly available or accessible via official databases such as data.gov.sg. Where relevant, data obtained from indirect sources (some slightly dated) will be used to supplement our passage. 3.2.1 Trial Waivers, Guilty Plea, and Plea Bargaining: Justice Short-Circuited? In-person negotiations 31 between the prosecution and the defence counsel for consensual case disposal in Singapore take place in two forums: the Criminal Case Management System (CCMS) or the Criminal Case Resolution (CCR) Program. 32 The CCMS allows Prosecution and Defence counsel to meet at an early stage to discuss the accused's case (including narrowing the issues in dispute or the merits of a guilty plea). Whereas, the CCR program was piloted late 2009 to reduce the rate of cracked trials 33 in the Singapore State Courts which at that point stood at a high of 43%. 34 The CCR program thus serves as another neutral forum for parties to negotiate consensual case disposal and resolution. CCR sessions are facilitated by a judge acting as a neutral party who can provide sentencing indications. 35 The judge as mediator will not give any assessment of the relative merits of the case but can comment on specific aspects of evidence and possible inferences or legal issues. 36 A criminal case typically goes through CCMS prior to CCR and may only be referred for CCR if there is a reasonable prospect of early resolution and if all parties voluntarily agree. 37 Both forums provide separate avenues for an accused person to plead guilty before their case proceeds to trial. It is important to recall that guilty pleas are not an unequivocal measure of an accused's guilt and their contrition. The decision to plead guilty may be linked to an accused's expectation of justice in the criminal system itself; where a country's high conviction rate may cause a defendant to believe that they have little chance of being 31 Note: Negotiations may also take the form of written representations to the Prosecution. 32 Laney Zhang, ÔPlea Bargaining: Singapore' (Library of Congress, 12 March 2020) ac cessed 1 March 2021. 33 A cracked trial occurs where the defendant elects to be tried but the case is resolved on the first day of trial or soon thereafter either by a plea of guilt by the defendant, or withdrawal of the charges by prosecution. A high incidence of cracked trials signify that Singapore's judicial resources are not being used optimally. 34 acquitted if they opted for their case to be heard. 38 In addition, sentence severity at trial is another significant factor that bears on the mind of a defendant weighing the costs and benefits of an early guilty plea. While no available statistics can confirm the conviction rate in the Singapore State Courts, data obtained from the Corrupt Practices Investigation Bureau (CPIB) in 2020 revealed that the conviction rate for cases charged by the CPIB is at 99 percent (i.e., almost all the suspects that were prosecuted were sentenced). 39 Despite the elite nature of corruption investigations, it may not be so far a stretch to infer that this conviction rate is maintained to some degree across other criminal cases heard in the State Courts. An accused person facing the Singapore judiciary may prefer to plead guilty if they have little confidence in their prospect of being acquitted. The controversial practice of plea bargaining has already been dealt with in numerous reports. 40 There is insufficient space in this paper to canvass the advantages and disadvantages of the process. However, some of its problems can be briefly identified. One critical disadvantage of the process, lack of parity, can be attributed to the absence of procedural safeguards during plea bargaining (e.g., evidence that may be inadmissible at trial can be brought up by the Prosecution during plea bargaining). 41 Plea bargaining also tends to exemplify prosecutorial powers 42 where initial overcharging of accused persons diminishes the reality of the bargain. 43 On balance, it is useful to consider the benefits of plea bargaining as promoting a more efficient and effective method of resolving cases without the need for trial. Plea bargaining also saves limited judicial resources and time. 43 It is argued elsewhere that the judge acting as a neutral overseer of the plea bargaining process under the CCR programme may bring more balance to the system. See: Koman (n 49).173 However, when put against a prevailing high conviction rate and an accused's expectation of justice delivery, the disadvantages seem to prevail. Independent legal advice is necessary for plea bargaining to operate in the safest environment and in the most equitable way possible. Access to such advice, however, is not guaranteed in the Singapore criminal process. 44 When legal advice is secured, it must also proceed minus the influence of investigating police officers who should be prevented from assuming a role in the provision of information to the unrepresented accused about the consequences of going to trial. 45 In any case, guilty pleas are the primary mode of case dispositions in most common law countries 46 and its propensity to undermine principles of due process cannot be ignored. The integrity of the pleabargaining process and its outcome requires transparency and an absence of unfair incentives to safeguard the defendant's rights in the criminal justice system. The number of cases disposed of via guilty pleas in the Singapore State Courts: There is no publicly accessible data determining the number of cases disposed of via guilty pleas in the State Courts. Suffice it to say that in or around 1996, Singapore's rate of guilty pleas is estimated as being above 90% 47 and more recent AGC figures from January to February 2017 revealed that 810 of the 851 convictions arose out of accused persons pleading guilty. 48 Arguably, with the emphasis placed on the CCR programme in pre-trial options, there will be stronger institutional inducement for accused persons to plead guilty. Sentencing indications from a judge will weigh heavily on an accused's decision to plead guilty. 49 better informed of the case against them, the defendant may choose to seek early resolution if they deem their prospect of succeeding at trial as weak. 50 State Courts' figures divulged that 75 out of a total of 119 cases (in between 2009 and 2012) that went through the CCR pilot process were resolved before trial and a total of 139 hearing days were saved. 51 In 2013, more than 80 percent of the cases referred for CCR were also resolved. 52 If nothing else, the image of the accused Ôhaving their day in court', testing the prosecution case, and having the presumption of innocence recognized in a procedural sense is not dominant in these administrative ''short circuits''. On the issue of involuntary guilty pleas submitted against an accused's wishes, the importance of defendants accessing effective counsel to eliminate wrongful convictions cannot be overstated. This is especially true since a defendant loses their right to appeal on a conviction once a plea of guilt had been entered into. 53 In the case of Yunami bin Abdul Hamid v PP, 54 the accused entered a plea of guilt for cannabis possession for trafficking. He was sentenced to 9 years' imprisonment and 6 strokes of the cane. Yunami later filed for criminal revision and his wrongful conviction was quashed by the High Court. The judge ordered a retrial, treating the accused's guilty plea as doubtful and observed that the accused faced overwhelming pressures that vitiated his ability to make a genuine and informed decision to plead guilty. Among these pressures was his inability to afford legal representation for trial. 55 Similarly, In Gao Hua v PP, 56 the High Court overturned her conviction on the basis that her plea of guilt was submitted following substantial pressures to plead guilty. The court also observed that a major factor that induced her sub- mission of a guilty plea was her worry that she would not be able to afford legal representation at trial. 57 While it is important to preserve the efficiency and limited resources of the courts, principles of due process and procedural fairness bearing on access to justice should outweigh administrative expedience. In a criminal trial, the resource imbalance between the state and an accused person means that both parties do not operate on an equal footing. The accused person is limited to their personal finances, legal aid, or pro bono assistance. 58 Determining State Court proceedings as trivial and less legally relevant based on limited penalties and high pre-trial resolutions is a one-sided evaluation. Attitudes preferring the disposal of cases going to trial by actively encouraging pleas of guilt should be scrutinised for its potential to impede justice. Any expectation for case resolution pre-trial because of the reported number of cracked trials in the State Courts should not operate to deny a defendant a fair trial. As McBarnet points out, ''offences and penalties may seem trivial from the outside but far from trivial from the perspective of the accused''. 59 The subjective measure of triviality and significance, if removed from lay participant experience and given over to legal professionals, administrators or machines/technology makes triviality another device for distancing vulnerable stakeholders from due process and participatory access. This analysis refrains from a critique of high plea-bargaining outcomes and resultant high conviction rates, in terms of Carlen and McBarnet's deeper interrogation of the failure of due process. Instead, by drawing attention to the dominance of administrative expedience, and recognising that many more cases are commenced at the State Courts level, the propensity to trivialize and regard these cases as legally irrelevant and hence offer less legality in proceedings is consequentially higher. As mentioned both by McBarnet and Carlen, redirecting any evaluation of the seriousness of the consequences of juridical determinations back to vulnerable stakeholders may not only invert this measure, but give genuine reasons for why due process protections and a greater array of justice options should be justified and not circumscribed through digitization. If determinations of triviality are pre-programmed into an otherwise unaccountable algorithm (e.g., one that in the future directs and mandates 57 Wong Siew Ming (n 69). the types of cases that are suited for CCR) then the Ôabsurdist' determinations of triviality are routinized and trapped away from any personal reconsiderations to the contrary. Introducing digitised mechanisms into this context has the potential to work on triviality and administrative efficiency as more important features for machine learning rather than ensuring due process protections which safeguards the integrity of the trial. In addition, the mystification surrounding the software, technology and data usage patterns of digitized justice will no doubt, further exclude meaningful appreciation and participation in a theatre of justice that takes on an automated, audio-visual unreality. If due process guarantees are insufficiently prioritised in the current plea process in the State Courts, the routinising of plea determinations through digital access, and the further distancing of the accused from professional human counsel could heighten the risk of abuse and lessen the likelihood of judicial accountability. A study of criminal trials in the State Courts between 1979 and 1980 revealed that defendants represented by counsel had a greater chance of being acquitted and a greater likelihood of receiving a less severe sentence than unrepresented defendants. 60 Nonetheless, save for capital punishment cases, accused persons in Singapore do not have a legal right to counsel paid for by the state. Hence, it is notable that a considerable number of defendants coming before the Singapore Courts are unrepresented. 61 Criminal Legal Aid in Singapore: 61 Our claim that a ''considerable number '' of accused persons are unrepresented is supported by three key pieces of information. First, save for capital cases, accused persons do not have a legal right to counsel. This means that accused persons are unlikely to be represented unless they are able to pay for their own counsel. Second, to qualify for legal aid, accused persons must satisfy both a means and merits test. Eligibility is therefore not automatic or guaranteed. Finally, as explained below, the legal aid scheme does not always support full legal representation at trial. 62 ÔCriminal Legal Aid Scheme' (Law Society Pro Bono Services) accessed 1 March 2021. individuals but to qualify for such legal assistance, accused persons need to satisfy both the Means and the Merits Test. 64 To pass the Means Test, the accused's disposable income and assets should not be more than S$10,000 over the last twelve months. 65 This figure should be read against the median salary in Singapore being $4500 (including employer's contribution) and average expenditure per household member as being over $1600. 66 Under the Merits Test, CLAS will assess whether the accused have reasonable grounds for defending their case. At this initial stage, the accused is expected to provide the facts of their case and the defence that they intend to rely on to the CLAS officer. If the case is deemed to be meritless, the accused's application would be denied. 67 To note, legal aid is not always fully subsidized and any payable legal fees would still depend on the individual's means test. 68 A co-payment amount may be required and any payment must be made within a stipulated deadline or defendants risk having their application rejected. 69 Out of pocket expenses such as the costs of obtaining police statements, reports, and transportation will also have to be borne by the accused. 70 It may take up to 3 months for a lawyer to be assigned. 71 The number of CLAS-assisted cases in the State Courts is not known due to the lack of official data. However, an estimated number of cases that qualify for CLAS can be determined from some available information. According to a press release in 2015, the CLAS aims to serve up to 6000 accused persons annually 72 but the scheme can only benefit up to 1000 defendants in terms of full legal representation at trial. 73 In 2020, the law society reported that it receives about 2400 applications for criminal legal aid every year and more than half receive full legal representation. 74 Accordingly, while exact figures are difficult to obtain, it is noteworthy that there are a considerable number of litigants who remain unrepresented despite a desire for legal representation. Since 90% of the cases are heard in the State Courts, 75 one can reasonably assume that most defendants are found unrepresented at this level. It is an open question whether defendants are unrepresented because their case is seen as trivial or less legally relevant. Yet, it should be recalled that unrepresented defendants are still expected to be held to the same standard of preparation and conduct as legal persons. 76 The PJP was set up to help accused persons in the plea negotiation process (i.e., not at trial). A lawyer can assist an accused person for a CCMS or CCR hearing to make representations or to engage in plea bargaining. Accused persons are charged a fixed fee of S$1000 for 3 h of legal services. 77 To qualify for the scheme, an accused's annual disposable income must not exceed S$12,000. 78 The PJP does not offer legal representation at trial but assistance is reserved only for cases where an accused person intends to plead guilty. From the above, it can be observed that most pro bono legal services are offered to accused persons at the legal information and advice stage. 79 The unavailability of legal representation beyond the initial plea negotiation stage may induce defendants to settle their case early. Pro bono legal advice appears to be the preferred form of legal assistance in Singapore but there is no space here to comment on the limitations of such schemes beyond highlighting that this does 74 Cara Wong, ÔNew Means Test Makes It Simpler to Apply for Criminal Legal Aid' The Straits Times (29 February 2020) accessed 8 March 2021. 83 ÔPrepare Your Mitigation Plea' (n 96). us with answers. One might be tempted to view the mitigation process, especially for State Court cases, as not requiring any legal expertise, advocacy or involving much law. 84 If it is any indication, when a case proceeds for CCR in Singapore, 85 it is observed that a session would only last for about half an hour and typically, not more than two CCR sessions would be conducted per case. 86 This short duration needs to be contrasted against the average length of a trial. Yet, perhaps more than in any other situation, in a process where the accused person is directed to offering a guilty plea rather than pressing her case, mitigation arguments require personalised and individualised professional representation. Even if the possible penalties seem minor against a wide range of punishments, what the accused considers to be significant in determining sentence may be of little relevance in the mind of the sentencer if not properly presented. The outcome of any such disconnection may diminish the possibility of leniency, particularly if the sentencing process moves to resemble some computerised calibration of like cases. On sentencing more generally, automated decision-making and predictive AI could potentially standardize and expedite the sentencing process. This would diminish the significance of or access to mitigation arguments. There has been much debate over recent decades about the possibility of predictive sentencing based on a computerized tabulation of exacerbating and mitigating factors. 87 However, along with standardization exists the potential that more sentencing decisions will treat the criminal process as automated thereby trivialising the individuality of punishment by denying the significance of discretionary decision-making. In addition, the influence of victim impact if this to be standardized would further diminish the possibility of tailoring each sentence to the circumstances of offending and the characteristics of each offender. Two issues arise here in the prospect of automated sentencing in circumstances where the technology learns that the sentence is trivial, and the determination of sentence is not so much dependent on law but rather on empirical correlation. The role of the judge as sentencer will be marginalized and further depersonalisation of justice will re-84 McBarnet (n 12). p.192 85 A defendant is allowed to plead guilty in the course of the CCR and a judge can take the plea and pass a sentence. 86 sult. In addition, available sentences will be routinised through algorithmic prediction rather than judicial discretion. Article 9(3) of the Singapore constitution provides that ''Where a person is arrested, he […], shall be allowed to consult and be defended by a legal practitioner of his choice.'' 88 The right to counsel is similarly represented under section 236 of the Criminal Procedure Code (The Code) which states definitively that ''every accused person before any court may of right be defended by an advocate.'' 89 This provision is narrower in scope since the right only applies when the accused is brought before a court. However, the right guaranteed under Article 9(3) is qualified because an accused person is not entitled to access their lawyer immediately upon arrest. In the case of Jasbir Singh v PP, the court established that an accused person is to only be granted access to a lawyer within a reasonable time of their arrest. 90 What is meant by ''reasonable time'' is to be balanced with the time required by the police to carry out their investigations effectively. 91 Police officers do not have to inform an accused of their right to counsel and can proceed to take an accused's statements 92 before they receive legal advice. 93 In Singapore, it is not uncommon that a suspect is denied 88 ÔConstitution of the Republic of Singapore' (SSO.AGC.GOV.SG) accessed 1 March 2021. 89 ÔCriminal Procedure Code (Chapter 68)' (n 67). 90 ÔIs It a Constitutional Right to Have Access to a Lawyer Immediately after Being Arrested?' (Gov.SG, November 2013) accessed 1 March 2021. 91 ÔIs It a Constitutional Right to Have Access to a Lawyer Immediately after Being Arrested?' (n 104). 92 Section 22(1) of the Criminal Procedure Code empowers the police to question and take statements from a potential suspect. 93 access to counsel until police investigation and statement extraction are completed. 94 Self-incrimination and admissibility of statements: Allegations of police abusing investigation powers have been reported in the press and through online channels. 95 Although section 22(2) of the Code contains a privilege against self-incrimination, the police are under no obligation to inform an accused person of this right. 96 The fact that a suspect is not told of this privilege prior to the making of a statement also has no impact on the admissibility of their statement at trial. 97 Statements recorded by the police are also rarely excluded by the courts for reasons of impropriety. 98 In 1995, it was reported in Parliament that from 1993 to April 1995, statements were ruled inadmissible in only 4 out of 166 cases tried in the High Court. There is no data available for cases heard in the State Courts 99 although it is imaginable that this figure will hover around a similarly low range. In any case, recalling that the majority of criminal cases are heard in the State Courts, the judiciary must be alert to possible impropriety on the part of investigating officers and to exercise its judicial discretion to exclude wrongfully obtained statements even if not strictly inadmissible. Judicial discretion over evidence admissibility is universally recognized as the essential filter for due process protections. This cautionary approach is relevant in countries like Singapore where the police treat the accused's initial statement as theirs and an accused's access to that statement is not of right. 100 Both McBarnet and Carlen were concerned about the disproportionate weight given to police narrative, where triviality and irrelevance may be determined early in the investigation process. With no access to a lawyer testing the propriety of police evidence, this disproportionate influence is confirmed. The advent of digital technologies in the statement-taking process may offer added protection but can create new problems for accused persons. For instance, while the use of audio-video recording can ensure that suspects are treated fairly in the interview process, its employment may interfere with a suspects' giving of evidence, offering another window (of opportunity) for negatively interpreting a suspects' demeanor, and evidential disputes surrounding such recordings including but not limited to the alteration/editing/leaking of such videotapes. If the admissibility of such recorded evidence was to be legitimated through technology, then digitising this process may further challenge access to justice and due process. On the other hand, there has been much support for the audio-visual recording of records of interviews for providing a Ôthird eye' onto an otherwise secretive and power-laden interaction. Looking forward, given the benefit of depersonalised and routine interpretation of the police narrative through digitized bias, the possibility of injustice may become more likely, yet more difficult to challenge. Developments in technology have had a positive influence on the openness, accuracy, and efficiency of information flow for the judiciary. 101 Technology has been harnessed in court systems for the purpose of case tracking, precedent analysis, and writing decisions. 102 Case tracking allows court administrators to monitor the progression of a case to determine whether procedural requirements have been satisfied. Greater access to legal sources enabled by technology assist judges in their decision-making process ensuring consistent decisions. Legal technology also supports judges with writing decisions, reducing the time required to deliver judgments. 103 provide audit trails to deter unauthorized access, eliminate incidences of misplaced files, and permit simultaneous access. 104 Singapore Courts are recognized as highly administratively efficient 105 and the country has leveraged the use of technological solutions for its courts as early on as the 1990s. 106 Online case management systems are employed in both the Supreme and State Courts 107 and the first technology court was launched in 1995. 108 Several other technology courts have since emerged. These courts are equipped with facilities such as video conferencing systems, computer-based recording transcription systems, litigation support systems, and integrated audio-visual systems. 109 110 Live cameras allow legal proceedings to proceed even without the physical attendance of parties or their lawyers. In 2019, the Singapore State Courts also introduced a real-time transcribing system. 111 The AI system is trained in court-specific vocabulary permitting it to transcribe English oral evidence in court hearings in real time, allowing judges and parties to review oral testimonies immediately. 112 Social distancing measures were implemented at the height of the pandemic in 2020 which halted the normal functioning of the Singapore Judiciary. 113 The country was however able to activate digi-tised court delivery relatively seamlessly due to its long-term emphasis on legal technology. This transition was enabled by Section 28 of the Covid-19 (Temporary Measures) Act. 114 To complement, the Singapore Courts also published a range of advisories designed to provide guidance to its court users. 115 Digitised legal support resources also remained available to members of the public. Existing legal aid applicants to the Legal Aid Bureau (LAB) were directed to submit their enquiries and upload their documents online, 116 while other eligible applicants may register their cases via the LAB's Applicant Portal and seek legal advice via video-consultation. 117 The Courts also remained open to members of the public 118 who could view ongoing trials via court monitors. 119 The brief coverage of technology cited here is not exhaustive but instructive in providing readers with an understanding of how technology is utilized by the Singapore judiciary. Court technology continues to evolve alongside new technological advancements and user demands. As witnessed in Singapore and across the globe, adaptation in times of crisis is a further push for digitised justice delivery. Digitalisation have the potential to radically change the traditional role of the judiciary and criminal justice service delivery at large. Across the board, technology has improved the efficiency of the courts and criminal process through better case management practices and court users' experience. These profits of digitisation notwithstanding, further advances in technologized justice need fundamentally to be measured against improving access to justice, recognising due process and procedural fairness. Earlier, we interrogated the operation of the ''two tiers'' framework in the Singapore judiciary and cautioned against the consequence of building on assumptions about triviality and legal irrelevance into the digital sphere. In this section, we elaborate on how digitised justice can worsen the access gap and McBarnet's existing ''two tiers'' through depersonalising and routinising justice service delivery. We examine this consequence of digitising practices in Singapore from the lens of Carlen's theatre of the absurd and her criticism of traditional courtroom features of space, time, and presentation as interfering with a defendant's (in)ability to participate in their own trial. We explain that the theatre and ''staging'' of the courtroom is reimagined in this digital dimension, but its isolating and exclusionary impact on the defendant is still keenly experienced. Justice is depersonalised when technology is introduced to substitute current court processes or services that presently involve communicating with a human actor or requiring an in-person human decision or experience no matter how mundane. The depersonalisation of justice can manifest on at least two distinct levels. First, when users seek to access court services but are directed to technologies designed to replace personalised interactivity and individualized information such as smart chatbots. While automation may lead to more certainty, predictability, and ''access'' in its broadest sense; individualized justice determinations become the casualty. ''Chatbot'' barriers to personalised access can become more resilient and harmful if the AI learns what type of enquiries are trivial or legally irrelevant and any consequences of its adverse decisions do not influence on the technology's mode of operation. Defendants directed towards these technologies and services are viewed by the technology as one entity, as a homogeneous group, whose cases are more administrative than it is legal. These court users could then potentially be denied access to legally relevant information (from a credible in-person source) that might prove crucial for their trial. If the peculiarities of their matter are universalized into one of a set of options, then the nuance of priority in crucial personal choices will be further distanced. Depersonalisation is similarly problematic when it surfaces in sentencing or mitigation. When traditional court processes (presently requiring human sensitivity and discretionary deliberation) are replicated by automation, option selection prioritizes efficiency to the exclusion of individualized deliberation. For appropriate cases, to recognize factual particularity, measured judicial discretion should be applied according to individual case circumstances. Technology or computer-based algorithms cannot be a substitute for judicial discretion and human judgment if justice demands subjective application. What can be said about depersonalised justice that widens the existing two tiers in Singapore? Our examination of emerging court technologies has shown that the process of depersonalisation tends to occur primarily at the State Courts level. The ideologies of triviality or legal irrelevance when programmed into automated choice options which have less regard for due process protections, create an environment of machine learning motivated by procedural efficiencies rather than procedural fairness. Indeed, where such court technologies have Ôlearned' triviality and irrelevance, their purpose would be to move minor offences out of the courts quicker -simplifying the decisions from the stage of arrest to sentencing, thereby enhancing time and cost savings -by effectively neutralising any impact of due process checks and balances. In exploring the phenological features of the Magistrate's Courts, Carlen observes that the mundane and conventional ways of magistrates' court staging (of time, space and rituals) are rehearsed then exploited to the defendant's detriment. Depersonalised justice and its technologies present the same paralysing effect in its depiction of all defendants and their ''trivial'' cases as being one in the same. ''Trivial'' cases are viewed as undeserving of the courtesy and majesty of the courts, thus withdrawing also fundamental principles of procedural fairness and due process safeguards. The following two instances of depersonalised justice in the Singapore Courts -the use of electronic guilty pleas and the creation and delivery of automated mitigation pleas -concretise these critical speculations. Carlen's observation of the Court's ''spacing, timing and presentation'' as operating to divorce the defendant from participation in their own trial can also be observed in this digital and depersonalised dimension. In spacing, these technologized solutions widen the Court's spatial dominance by subjecting the defendant to an automated process where access to legal information and assistance becomes even more diluted and difficult to access. A defendant's digital illiteracy can also further exclude them from any engagement. This breakdown of two-way communication is reminiscent of the poor acoustics, placing and spacing of the defendant in the magistrate's courtroom. Moving on to time, these automated processes enable the quicker handling and disposal of ''trivial'' cases to preserve judicial and professional resources. Yet, in this exercise, the defendant is left stranded, with only their time squandered away if they are unfamiliar with the digital ways of seeing and doing. In the same way how the timing of magistrates' justice is monopolised by the police, control of the proceedings now lies in the hands of machines that dictate and reduce these defendants' options to depersonalised choices. Finally, with presentation, defendants receive and is reliant on instructions given to them by the machine. The machine replaces the court room, and the defendant is subject to its limited commands. The presentation style of the defendant (in Court) is lost and, in any case, is now irrelevant in this digitised dimension. We look at two examples here. Electronic guilty pleas: In Singapore, it is possible to plead guilty by electronic means for some minor traffic offences when an offer of composition has expired. 120 This plea of guilt may be entered into at an AXS automatic bill-payment station, and the offender can plead guilty to the offence without needing to attend court. 121 In this expedited process, the accused loses out on the opportunity to seek advice from a lawyer or court personnel; sacrificing any opportunity for the accused to appreciate the nature of their plea and punishment, even when the offence in question carries a possible imprisonment term. 122 Machines cannot provide sufficiently personalised information to equip an accused with enough knowledge to make their judgment but are designed to reduce individual choice to a Ôyes/no' binary. The accused's experience in the conventional plea-bargaining process is replaced with a push of a ''guilty'' button. The plea-bargaining process and public justice service delivery is depersonalised because the lawyer, the judge, the live court, and its visitors are now absent features in this digitised process. Digital innovations like electronic guilty pleas are recommended to speed up the sentencing of minor offences. Rational choice theory suggests that defendants would likely pick the option that is convenient and less stressful. Thus, the likelihood of an accused person preferring to go to court is sharply reduced when pleading electronically is presented as an option. As such, justice deliberations are not only depersonalised here but the decision of the accused as to the preferred style of justice service delivery is constrained by convenience over contestation. Electronic mitigation pleas: The PJP offers an e-guided mitigation plea option 123 which is a self-help system where accused persons draft their own mitigation plea via the Community Justice Courts's (CJC) Automated Court Documents Assembly (ACDA) system. The plea is later reviewed by a lawyer at a fixed fee during a 1-hour legal consultation. Access to this e-mitigation platform is only available to accused persons employing PJP's services. 124 The public has no general access. However, based on information that is publicly available, there appears to be no restrictions on the type of cases that can utilise this e-guided option. 125 The CJC had previously indicated that a ''mother [had] used the system for her mitigation plea to the judge for shoplifting.'' 126 The outcome of the case is not known but it 122 The schedule of the act stipulates that an accused may plead guilty by electronic means to an offence that is punishable by imprisonment not exceeding 3 months. 123 ÔPrimary Justice Project (PJP)' (n 77). 124 ÔPrepare Your Mitigation Plea' (n 96). 125 ÔPrepare Your Mitigation Plea' (n 96). Note: limitations may apply once access to the system is granted or after an accused's consultation with the PJP. If this is the case in practice, the e-mitigation option will only be available for certain cases. However, from our knowledge and as reflected on the Singapore Courts websitethere are no outright restrictions (on the type of cases that can rely on this option) as reflected on its ''prepare your mitigation plea'' page. 126 Leonard Lee, ÔInnovation in the Pro Bono Scene: An Interview with Leonard Lee' [2018] Singapore Comparative Law Review 132. is notable that there is a mandatory imprisonment term (which may extend to 7 years) for shoplifting offences in Singapore. 127 E-mitigation pleas marginalize an accused's due process protections and any consequent expectations for a fair hearing. How is the accused to know what mitigating factors to input in their plea and whether the factors have been correctly applied or conveyed? On depersonalised justice in particular, such interfaces fail to convey to the accused the significance of adequate legal representation. These platforms may tempt the accused into believing that the quality of legal ''advice'' received is of a sufficient standard as that offered by a human lawyer. Yet, there are always nuances in the facts of any alleged offence and in any individual's particular circumstance. Such variations will not be adequately captured by standardized questions and answers in automated formats. A 1-h legal review of the plea by a legal professional may not consider these peculiarities. The effectiveness and ''quality'' of such standardized mitigation pleas in the mind of the judge as a sentencer also merits reflection. Minor offenders are also more likely to rely on automated mitigation pleas than more serious offenders. In this way, there is a Ôtriviality and irrelevance' self-selection enhanced through automation. This normalization of depersonalised deliberation would inevitably exacerbate McBarnet's two tiers divide (now between automation and personalised deliberation) as more offenders facing the State Courts go to the machine instead of seek personalised legal assistance. In this manner, depersonalised justice deliberations push the triviality/irrelevance conclusion down to the accused through automated exigencies. Justice is routinised when arguably simple or straightforward court processes are automated through digitalisation and as a consequence make Ôyes/no' automated binaries the default. The routinisation of justice delivery goes a step further away from due process protections than depersonalising justice because not only is the human ''divorced from the loop'' but the recipient of automated justice is excluded, or in McBarnet's terms, ''silenced'' from participation in their own 127 ÔPenal Code' (Singapore Statutes Online, Section 380 (Theft in Dwelling House)) accessed 10 November 2021. Note: There is also the offence of theft under section 379 punishable with imprisonment for a term which may extend to 3 years, or with fine, or with both. proceeding. Justice is routinised when a defendant's issues are set up as mundane and less deserving of the Court's procedural pedantics. Digital exclusion in this sense occurs because the user's interaction with the court system is limited or eliminated, to distance human contestation from predictable coincidences and causation trails. Predictability from a set of pre-determined and routine options is the essence of administrative frugality. Routinised justice can be deliberately induced (i.e., when technology is installed to achieve a particular outcome in justice delivery from a utilitarian cost-benefit frame) but may also transpire as an unintended consequence of digitalisation (i.e., neglecting to consider the consequent impact that digitalisation may have on an accused person). Routinising justice is a quest to deny idiosyncrasy in favour of categorization and prediction. It links to depersonalising in so far as removing discretion from decision processes enables an ordering of choice options that force the individual requirements of the justice service user into pre-determined availabilities not based on user need (because these needs have already been marginalized as trivial or irrelevant) but more for efficiency and cost-benefit in administration. Justice service delivery so managed will be more streamlined and as such, appear available to more users. However, digging a little deeper past delivery administration to look into service quality measures, and the nature of justice on offer may appear fundamentally changed. True it is that many of the matters so processed are minor on an objective scale of seriousness, but particularly in Singapore where criminal punishments even at the lower end can be substantial, as McBarnet comments, the importance of the outcome to the accused person should not be so diminished. Routinised justice, would be for Carlen another way of distancing the accused from their trial. The space, time and presentation (or so-called hierarchy) between the insider Court and the outsider defendant becomes more pronounced and significant in these digitised models where the defendant is separated not merely by the physical structures of the court room layout (and courtroom etiquette) but also by digital formalities that produce novel coercive structures that ignore individual vulnerability. Delivery and Fluency of e-hearings and fair trials: Virtual court hearings have been employed to meet justice demands in the past year 128 and will likely form a more common part of Singapore's digital transformation in legal service delivery. The use of video hearings restricts the fluency of dialogues at trial especially with the presentation of evidence. 129 From a visual perspective, the sharing of legal documents on platforms such as Zoom can relegate the camera feeds of parties to the side or top of the screen where the camera feed of the person giving evidence is minimized, often obstructing the Court's view of the individual. 130 The integrity of witnesses and their giving of evidence is also challenged by the use of virtual hearings through the limitation of non-verbal cues which judges often consider revealing. Digital trials can therefore render cross-examination more difficult due to the lack of a ''frontal encounter''. 131 Video-conferencing software may over or under-emphasize certain interactive gestures, impacting the accurate assessment of the credibility of witnesses during cross-examination. 132 Technical failures resulting from the use of video hearings are also not uncommon 133 and nonverbal cues such as postures, gestures, and eye contact may also be distorted by interruptive video streams or the positioning of the video frame. 134 In this online format, the defendant's participation is routinised as necessary and as the preferred option (the defendant not having a say on the format the trial takes) since the system prioritises digital hearings as a health necessity. Carlen's ''theatre of the absurd'' is exacerbated by digitalisation and routinised through the normalization and mundanity that technological distancing and automation suggest. The accused becomes a bystander in their own trial as matters proceed without compensating for how digitalisation interferes with their physical participation. Such participation is now additionally dependent on their comfort and familiarity with the use of the technology. A defendant in this technologized terrain is not merely separated by conventional professional presentation, court space, and compacted hearing time components but the added digital infrastructure and device(s) contribute another layer of distancing between the accused and the court. Technology adds a further realm of professional dependency which the accused must penetrate. The theatre is no longer played out in the courtroom but on the computer screen and justice becomes a televised performance with all the unreality that pertains. The loss of the ''majesty of the courts'' and what it represents for the pursuit of justice should also not be overlooked as a natural consequence of digitised justice. It is argued elsewhere that the transition to online and automated justice disregards the atmosphere and solemnity of in-person trials including the presence of judicial authority. 135 But more than courtroom presentation and traditions, the majesty of the courts is arguably linked to the law's safeguarding of procedural fairness and due process protections. Now it is the zoom host that takes control. We have earlier observed that this physical solemnity was already fast fading in the lower courts where there is a lack of investment in certain human-centered processes because of the desire to reduce cracked trials. The discussion of depersonalised justice above has also shown how certain court processes (particularly in the lower courts) when digitised compromise the user's experience and interaction with the judiciary in situations where otherwise an in-person trial would eventuate. Digital access and cybersecurity issues: Digital justice can be said to generate benefits such as enhancing judicial transparency, improving the effectiveness, efficiency, and expediency of case management, and facilitating increased access to quality legal information and processes for professionals, administrators and those who know how to traverse the technology. Notwithstanding these advantages, two adverse considerations are obvious when thinking about the routinisation of justice: namely, who is left behind and what is at stake. Routinised justice seeks to produce standardized processes that save judicial resources, costs, and time. Motivated by this frame, technologies that are deployed tend to ignore vulnerable populations and 135 Yoong (n 128). their unique needs in accessing justice. Routinised justice produced by digital solutions disregards independent user characteristics. Certain court users, particularly, litigants-in-person, may lack access to, or experience difficulty participating in remote hearings. 136 Parties may also lack the digital knowledge necessary to navigate case management systems or enter court applications online. Digitised legal aid delivery is another service that could exclude persons who are less digitally equipped in terms of their access to digital tools or knowhows. Financial constraints similarly influence who has the ability or means to access digital solutions. These circumstances, whether independently or cumulatively encountered, produce impediments to access to justice that depreciate the quality of justice service delivery. Digital exclusion posed by these technological projects surfaces in both the lower and higher courts. The exclusionary capacities of any proposed court technology will not discriminate against persons who commit a minor as opposed to a more serious offence but are more likely to be encountered by those that the system deems less worthy of personalised justice. In that sense, whether one is digitally excluded (or included) from digital participation in the justice sector is ultimately dependent on an individual's circumstances combined with a Ôtwo tiers' devaluation of the significance of the matter in question. Obviously, there is a greater propensity to routinise and digitise processes that are determined as involving minor offences, with the intention of diverting these away from human determinations and limited resources. Another unintended consequence of digitalisation and routinised justice is the emergence of cybersecurity threats. In the pursuit for greater judicial efficiency, there is a desire to collect, retain, utilize, and repurpose large amounts of personal data in digital formats. Routinization injects ''ordinariness'' and ''conventionality'' in ways of doing so that important questions are either ignored or forgotten. This enables intrusive data collection and access practices to be superimposed in the same way that mundane ways of organising and communicating in the magistrates courts are exploited to the defendant's detriment. For example: • when data is collected for its own sake (rather than for a specified justice purpose benefitting the court user), • or stored in ways that compromise individual data subject privacy (replicating already-familiar processes and systems for convenience), • or when good digital hygiene practices are bypassed (because certain practices have become so commonplace not to invite critical review) In these instances, data integrity and security are at issue. If data is not appropriately safeguarded, defendants are exposed to common cybersecurity risks. This could result in highly sensitive information being leaked which may have damaging consequences on individual privacy. 137 With the move towards online case management systems and online dispute resolution platforms, these threats are becoming even more commonplace at the institutional level. In 2018, 223 State Courts online case files were retrieved without authorization due to a loophole in the ICMS system. 138 Data accessed from the files included the names of the accused, their NRIC, addresses, gender, and other information about the offence. 139 Such incidences are of course not restricted to the lower tier of justice. However, recognising that the majority of cases are heard at the State Courts level, more data is evidently at stake here. In proposing new possibilities for justice delivery through a more digitised format, Zorzi reminds us: This is the occasion for justice systems around the world to adopt a peoplecentered approach to justice, to remove barriers to innovation and technologies that can revolutionize the way in which justice is delivered, to establish open, inclusive, fair, and accountable justice systems that work closely with 137 Jasmine Ng, Andrew Wong and Jia Qing Yap, ÔLegal Tech-Ing Our Way to Justice' (Law Tech Asia) accessed 24 February 2021. 138 Lydia Lam, Ô223 State Courts Online Case Files Accessed without Authorisation Due to Loophole' Channel News Asia (28 November 2018) accessed 25 February 2021. 139 Lam (n 154). health and other sectors, that make people their partners and inspire trust in the institutions. 140 Therefore, the challenge in digitising justice is not so much for achieving cost-savings and administrative efficiency, which is generally accepted, but to employ informed and flexible technologies that are people-centred and value inclusion. If these were the priorities for digitising justice, then automation and technological innovation would work to redress the two-tier divide, or at least counter its excesses in the subordinate courts. If digitising justice is to be part of the solution and not just the problem in terms of reducing procedurally protected access and exacerbating the two tiers divide, then ongoing, advocates of this revolution need to: • Selectively identify examples of where the delivery of justice currently does not meet access and due process measures because of assumptions about triviality and irrelevance. Building on these examples, policy makers should address the negative impact of digitising that exemplifies and compounds what are already problematic institutions and processes in justice service delivery open for further depersonalising and routinising. Once identified, countermeasures need to be developed within digitised justice learning and programming to recognize universal access and due process principles. In particular, as Zorzi proposes, openness, inclusion, fairness and accountability sit well with the general aspirations for ethical AI. Digitised justice should be identifying rather than replicating ideologies of triviality, irrelevance, routinising and depersonalising that in pre-digitised frames exemplify the two tiers both in normative justice imaginings and practical day-today service delivery. The learning criteria for automated justice technologies and software is to operate beyond the ideological and functional failings that the Ôtwo tiers' exposes. • To achieve Zorzi's invocation for a more people-centred approach to justice through digitising, the same examples should be tested against ways that depersonalising and routinising can be minimized through digital solutions that have not learnt a version of justice service delivery in the context of triviality and irrelevance. The disturbing reality in what has been revealed about the Ôtwo tiersÔ Singapore style pre-digitising will be exacerbated when digitising 140 Zorzi (n 1). justice works from triviality/irrelevance and further depersonalises and routinises justice service delivery. The cynic might say that access to justice and the assurances of due process and procedural fairness are already so distant for most accused. As such, digitising is only more of the same. In refuting this concession, there is nothing inherent in digitising through automation and technology that must confirm and continue the worst of what is present. Even if the driving motivation for digitised justice is administrative efficiency, streamlined case-management and more purposeful data management, these cannot only be achieved through routinising and depersonalising. As technological aids to alternative dispute resolution reveal, digital technology can make personalised and individualised human decision-making a reality. There has not been sufficient time to fully interrogate the prevailing outcome of two tiers and theatre of the absurd, exclusion of the stakeholder for whom justice is meant to be delivered. If digitised justice operates on the understanding that not only are the matters in question, but also the accused persons to whom they relate, irrelevant and trivial, then no matter how procedurally fair the algorithm or the tech may be, the two tiers will prevail. Identification, recognition, and location of the accused in the digital space of the courtroom and criminal process will not be sustainable unless and until the accused is empowered in the management of their data on which digital transformation depends. Digital self-determination placing the data-subject, their autonomy and communal duty at the forefront of data management is a way in which even the most marginalised accused can be empowered to operate and benefit from the data at the centre of their contestation. If the digitising project is premised on individualised access and empowering due process, then digital self-determination can better balance the power differentials that presently divide the litigant from fair trial. Through such a commitment to people-centred digital transformation, technologising trial justice will offer new and equitable data spaces not pre-determined by Ôtwo tiers' discrimination and alienation. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. )