Skip to content
laws.sgby aturio
Grounds publishedDecided 10 Apr 2026
Contents

General Division of the High Court

A fine, not jail, was enough for a low-blame receiver of confidential information

Tran Thi Tien v Public Prosecutor

Feedback & support

[2026] SGHC 80Case no. Magistrate’s Appeal No 9109 of 2025

The decision, in one line

Not legal advice

A fine, not jail, was enough for a low-blame receiver of confidential information

Summary by laws.sgParagraphs [175], [182]

There is no fuller plain-English summary of this judgment yet.

Ask about this judgment

Laws this judgment relies on

Grounds of decision

Official text from eLitigation

  1. Introduction and procedural background

  2. [1] This appeal concerned the sentencing of an offender convicted of unauthorised receipt of confidential information under s 5(2) of the Official Secrets Act (Cap 213, 2012 Rev Ed) (“OSA”). The key issue which arose for consideration in this appeal concerned the appropriate sentencing approach to be adopted for such offences.

  3. [2] By way of background, the appellant, Tran Thi Tien (“Appellant”), pleaded guilty in the District Court to one charge under s 5(2) of the OSA, punishable under s 17(2) of the OSA, of receiving information, having reasonable grounds to believe, at the time when she received it, that the said information was communicated to her in contravention of the OSA. The District Judge (“DJ”) convicted the Appellant of the said charge and sentenced her to one week’s imprisonment. The DJ’s grounds of decision are set out in Public Prosecutor v Tran Thi Tien [2025] SGDC 191 (“GD”).

  4. [3] The Appellant appealed against her sentence and was on bail pending the disposal of her appeal. The crux of her appeal was her contention that a custodial sentence was not warranted in her case, whereas the Prosecution contended that it was. The hearing, which was originally fixed for 3 November 2025, was adjourned to allow both the Appellant and the Prosecution to put in further submissions addressing my queries as to what the appropriate sentencing approach should be for s 5(2) OSA offences, having regard to the dearth of reported sentencing decisions for such offences.

  5. [4] Having considered parties’ written and oral submissions, including their further written submissions, I allowed the Appellant’s appeal against sentence, and substituted a fine of S$1,500 for the sentence of one week’s imprisonment. These are my full written grounds of decision which include, in particular, my findings as to what the appropriate sentencing approach should be for offences of unauthorised receipt of confidential information under s 5(2) of the OSA.

  6. [5] The version of the OSA considered in this case is the version in force at the time when the Appellant committed the offence (ie, 10 December 2020). However, there is no difference between the sections material to the present case which were in force as of 10 December 2020 (ie, ss 5(2) and 17(2) of the OSA), and the corresponding provisions which are currently in force. I also highlight that in these written grounds of decision, I use the term “confidential information” as a convenient shorthand for referring to “information communicated in contravention of the OSA”, and this term does not relate to any particular sensitivity or security classification. The term “confidential information” was used in a broadly similar manner in a speech on 10 January 2011 by the Minister for Home Affairs, Mr K Shanmugam, in response to a Parliamentary Question regarding the WikiLeaks incident (in which classified diplomatic cables had been published online). I return to the content of this speech at [52] below.

  7. [6] I first set out the undisputed facts in this case, based on the Statement of Facts (“SOF”) to which the Appellant admitted in pleading guilty.

  8. Background

  9. Undisputed facts

  10. [7] The co-accused in this matter is one Chan Zhiyao, who was, at the material time, a Senior Investigation Officer attached to the Special Victims Unit of Bedok Division, Singapore Police Force (“SPF”). The co-accused joined the SPF in 2003.

  11. [8] The SPF employs a confidential computer database system known as the CRIMES II system, on which police officers may conduct screenings of individuals to obtain information about them. This includes information about previous cases of police investigations involving the individuals being screened. Police officers may log onto the CRIMES II system using their personal username and password. By virtue of his position as a Senior Investigation Officer of the SPF, the co-accused had access to the CRIMES II system.

  12. [9] Sometime before 2 December 2020, the co-accused was assigned to investigate a case of child abuse involving the Appellant and her ex-husband, one Wong Peiliang (“Appellant’s ex-husband”).

  13. [10] In the course of these investigations, the Appellant informed the co-accused that she and her ex-husband were undergoing divorce proceedings.

  14. [11] On 9 December 2020, the Appellant informed the co-accused that sometime around end-2016, her ex-husband had stolen a person’s handphone at Tan Tock Seng Hospital (“Theft Case”). The Appellant asked the co-accused to provide her with information on the investigations against her ex-husband in relation to the Theft Case, so that she could relay the same to her divorce lawyer.

  15. [12] Sometime between 5.40pm on 9 December 2020, and 8.24pm on 10 December 2020, the co-accused conducted an unauthorised screening of the Appellant’s ex-husband’s name using the CRIMES II system, to procure information relating to the Theft Case. The information regarding the Theft Case was not related to the child abuse case that the co-accused was investigating. Through this unauthorised screening, the co-accused obtained the police report number (“Report Number”) pertaining to the Theft Case.

  16. [13] On 10 December 2020, at 8.24pm, the co-accused sent the Appellant the Report Number via WhatsApp. The Appellant knew that the co-accused had obtained the Report Number owing to his position as a Senior Investigation Officer who held office under the SPF. The Appellant also knew that the Report Number was unrelated to the child abuse case that the co-accused was investigating. She therefore had reasonable grounds to believe that the co-accused was not authorised to communicate the said Report Number to her, and that it was not the co-accused’s duty to communicate the said Report Number to her.

  17. [14] As the Appellant received the Report Number, having reasonable grounds to believe, at the time of receipt, that this information was communicated to her in contravention of the OSA, the elements of an offence under s 5(2) of the OSA (punishable under s 17(2)) were made out.

  18. [15] The above offence was discovered by investigators when they were conducting unrelated investigations into the co-accused. The investigations also revealed that the Appellant and the co-accused had entered into a sexual relationship on or around 25 May 2022.

  19. [16] Per the SOF, there was another incident following the above offence. This involved a WhatsApp conversation between 8 September 2022 and 9 September 2022, in which the Appellant asked the co-accused whether a shopfront that she intended to rent had previously been “checked” by the Police. On that occasion, the co-accused refused to provide the information, stating that it was confidential.

  20. [17] The SOF further recounted that the Appellant gave the co-accused a loan of S$2,000 on 5 January 2023. The Appellant was subsequently arrested on 17 October 2023 and released on bail on the same day.

  21. Decision below

  22. The sentencing considerations and factors considered by the DJ

  23. [18] I next summarise the sentencing considerations and factors considered by the learned DJ in sentencing the Appellant.

  24. [19] To begin with, as the DJ noted, the OSA prohibits the unauthorised disclosure and unauthorised receipt of confidential information, given that such acts “threaten the integrity of State information” and might have “potentially serious ramifications”. This means that the principal sentencing consideration for OSA offences is general deterrence of both public officers and external parties. In his GD, the DJ explained that the offence of illegitimately procuring confidential information from the police force is “essentially an offence against a public institution”, and that offences against public institutions generally warrant general deterrence (citing Public Prosecutor v Law Aik Meng [2007] 2 SLR(R) 814 (“Law Aik Meng”) at [24(a)]). That said, the DJ acknowledged the trite principle that “deterrence must be tempered by proportionality”.

  25. [20] Noting the paucity of reported sentencing decisions on OSA offences, the DJ took reference from corruption offences that similarly involved undermining of public institutions. Citing Goh Ngak Eng v Public Prosecutor [2023] 4 SLR 1385 (“Goh Ngak Eng”), the DJ considered the following non-exhaustive offence-specific factors to be relevant in assessing an offender’s culpability and the harm posed by their offence:

  26. [21] As for offender-specific factors, the DJ considered the following non-exhaustive factors (citing Public Prosecutor v Wong Chee Meng [2020] 5 SLR 807 (“Wong Chee Meng”)):

  27. [22] Finally, the DJ held that the principle of parity in sentencing applied in the present case. Citing Chong Han Rui v Public Prosecutor [2016] SGHC 25 (“Chong Han Rui”) at [47], the DJ held that the relevant question was whether the public would perceive that an offender who was party to a common criminal enterprise had suffered injustice by reason of a disparity in sentence vis-à-vis a co-offender who was also party to the same common criminal enterprise. While the principle of parity had yet to be applied in reported decisions on OSA offences, the DJ found it helpful to consider corruption offences which similarly involved conduct undermining public institutions. He noted that in the context of such corruption offences, the general principle was that the giver and recipient of gratification were equally culpable, although the sentencing court must still assess the culpability of each offender and their unique circumstances.

  28. [23] Having set out the above sentencing considerations and factors, the DJ applied them to the Appellant’s case in the following manner.

  29. [24] Beginning with the Appellant’s culpability, the DJ first considered Public Prosecutor v Phua Keng Tong [1985–1986] SLR(R) 545 (“Phua Keng Tong”). In that case, the receiving party (Phua) claimed trial to a charge under s 5(2) of the Official Secrets Act (Cap 233, 1970 Rev Ed). He was acquitted by the trial judge, but the prosecution’s appeal against the acquittal was allowed by the High Court, which convicted him of the charge and sentenced him to a fine of S$1,500. The DJ distinguished Phua Keng Tong from the present case on the basis that the Appellant had “specifically solicited the information from [the co-accused]”, whereas in Phua Keng Tong, the receiving party had not asked for confidential information.

  30. [25] The DJ held, moreover, that the Appellant’s culpability was “heightened” by her “ill intent” and “malice”, in that she had sought confidential information which she intended to “weaponi[s]e in divorce proceedings against her ex-husband”. In contrast, according to the DJ, the disclosing party in Phua Keng Tong had offended due to “a lapse of judgment, absent any malicious intent”. The fact that the Appellant eventually did not use the Report Number was held to be a “neutral factor”, particularly since the “potential for misuse” [emphasis added] of the information remained, as she “could still use that information against her ex-husband in the future”.

  31. [26] Based on the above reasoning, the DJ concluded that the Appellant’s culpability was “materially higher” than that of Phua in Phua Keng Tong.

  32. [27] As for the harm caused by the Appellant’s offending, this was held to be low but not negligible, and in the DJ’s view, “slightly higher” than the harm occasioned in Phua Keng Tong. He found that the very fact that confidential information (ie, the Report Number) had been leaked from the CRIMES II system to the Appellant meant that actual prejudice had been caused to the SPF’s interest as well as the ex-husband’s interest. Also relevant, in his view, was the potential for future misuse of the information against the Appellant’s ex-husband.

  33. [28] In respect of the parity principle, the DJ held that this applied to the present case because the Appellant and the co-accused were “participants of a common criminal enterprise”, in that the former had “solicited the information, and [the latter] provided it”. He found that the co-accused, being well-positioned to refuse the request for confidential information, was “significantly more culpable” than the Appellant. Since the co-accused had been sentenced to two months’ imprisonment for his OSA offence, the DJ held that a sentence of one week’s imprisonment for the Appellant would be commensurate with her reduced culpability.

  34. [29] In arriving at this sentence, the DJ took into account the Appellant’s early plea of guilt, which entitled her to a discount in sentence. On the other hand, he also took into account her subsequent attempt “some two years later” to find out from the co-accused whether a shopfront that she intended to rent had been previously “checked” by the police. According to the DJ, this subsequent incident “reinforced [his] view that the prior offence was not a momentary lapse of judgment”.

  35. Issues to be determined

  36. [30] The main issue in contention in this appeal was whether the DJ erred in imposing a sentence of one week’s imprisonment instead of a fine. As I noted earlier, this gave rise to an anterior question as to the appropriate sentencing approach for offences of unauthorised receipt of confidential information under s 5(2) of the OSA.

  37. [31] Before I address the above issues, I should note for completeness that while the punishment provision in this case (s 17(2) of the OSA) states that a person who is guilty of an offence under s 5 of the OSA “shall be liable on conviction before a District Court to a fine not exceeding $2,000 and to imprisonment for a term not exceeding 2 years” [emphasis added], both parties (as well as the DJ) appear to have understood s 17(2) of the OSA to confer a discretion on the sentencing court to impose sentences of fines and imprisonment terms as alternatives. On appeal, both parties confirmed that this was their common understanding of how s 17(2) should be construed. I accepted their common position. As See Kee Oon JC (as he then was) observed in Sim Wen Yi Ernest v Public Prosecutor [2016] 5 SLR 207 (at [52]), the words “shall be liable” do not have any prima facie obligatory or mandatory connotation, and have been generally viewed as conferring a discretion, while the word “and” may, in some cases, be used in a disjunctive sense. The parties’ common position – that s 17(2) gives the sentencing court the discretion to impose sentences of fines and imprisonment as alternatives – is consistent with the position taken by the High Court in Phua Keng Tong, which imposed a fine of S$1,500 on the receiving party Phua after convicting him of the s 5(2) charge.

  38. The principles governing appeals against sentence

  39. [32] Parties were also agreed as to the principles governing appeals against sentence. Statutorily, s 390(1)(c) of the Criminal Procedure Code 2010 (2020 Rev Ed) (“CPC”) provides that in an appeal against sentence, an appellate court may “reduce or enhance the sentence, or alter the nature of the sentence”. Further, an appellate court may reverse or set aside a trial court’s sentence that is found to be “wrong in law or against the weight of the evidence”, or “manifestly excessive or manifestly inadequate in all the circumstances of the case”: s 394 of the CPC. An appellate court “will not ordinarily disturb the sentence imposed by the trial court except where it is satisfied that: (a) the trial judge erred with respect to the proper factual basis for sentencing; (b) the trial judge failed to appreciate the materials placed before him; (c) the sentence was wrong in principle; or (d) the sentence was manifestly excessive or manifestly inadequate, as the case may be”: Public Prosecutor v UI [2008] 4 SLR(R) 500 at [12]. A “manifestly inadequate” or “manifestly excessive” sentence is one which is “unjustly lenient or severe, as the case may be, and requires substantial alterations rather than minute corrections to remedy the injustice”: Public Prosecutor v Siew Boon Loong [2005] 1 SLR(R) 611 at [22].

  40. Issue 1: Appropriate sentencing framework for s 5(2) OSA offences

  41. [33] I now address the issue of the appropriate sentencing approach for offences under s 5(2) of the OSA.

  42. [34] Parties’ original submissions focused primarily on dissecting the DJ’s GD, and on arguing about whether he was justified in drawing certain inferences from the evidence before him and in concluding that the inferences so drawn warranted a custodial sentence. However, as I pointed out to parties when asking for further submissions, this exercise can only be carried out in a meaningful manner if all parties first share a clear understanding of the key sentencing considerations applicable to offences under s 5(2) of the OSA, as well as the specific factors to be considered by the sentencing court in calibrating the seriousness of a s 5(2) OSA offence. In particular, being able to coherently articulate the relevant sentencing considerations and factors would provide us with a principled basis for determining when the custodial threshold is crossed for s 5(2) OSA offences and whether the threshold has been crossed in a particular case. This, in turn, should (hopefully) bring about greater consistency and transparency in the sentencing of s 5(2) OSA offenders in future cases.

  43. [35] In endeavouring to articulate the sentencing considerations and factors relevant to s 5(2) OSA offences, parties were agreed that given the dearth of reported sentencing decisions for such offences, it was necessary to approach the matter from first principles. At the outset, the Appellant did not seriously disagree with the Prosecution’s submission that the predominant sentencing consideration for offences under s 5(2) of the OSA should be general deterrence. With this guiding principle in mind, parties were also broadly agreed that given the wide range of different circumstances in which s 5(2) OSA offences could be committed, it would not be feasible to classify the seriousness of an offence by reference to a set of “principal factual elements” – as has been done, for example, in relation to vice-related offences (see, for example, Poh Boon Kiat v Public Prosecutor [2014] 4 SLR 892 (“Poh Boon Kiat”), as discussed by the Court of Appeal in Ng Kean Meng Terence v Public Prosecutor [2017] 2 SLR 449 (“Terence Ng”) at [33]–[34]). Instead, both parties proposed adopting – with some adjustments – the sentencing framework established in Logachev Vladislav v Public Prosecutor [2018] 4 SLR 609 (“Logachev”) for offenders who claim trial to offences punishable under s 172A(2) of the Casino Control Act (Cap 33A, 2007 Rev Ed) (“CCA”).

  44. [36] The Logachev framework (Logachev at [75]–[84]) may be summarised as follows:

  45. (a) The first step of the framework requires the sentencing court to have regard to relevant offence-specific factors, in order to identify the level of harm caused by the offence and the level of the offender’s culpability. The level of harm caused may be categorised into three levels (ie, slight, moderate or severe), and so too may the offender’s level of culpability (ie, low, medium or high);

  46. (b) Second, the court should identify the applicable indicative sentencing range with reference to these two factors, ie, the level of harm caused by the offence and the level of the offender’s culpability;

  47. (c) Third, the court should identify the appropriate starting point within the indicative sentencing range. This is to be done with regard, once again, to the level of harm caused by the offence and the level of the offender’s culpability. While this step will engage the same offence-specific factors as those considered at the first step, it is not an instance of double-counting: rather, this step involves the court granulating the case so as to derive a sense of the appropriate starting point in that particular case;

  48. (d) Fourth, after identifying the appropriate starting point within the indicative sentencing range, the court may adjust this starting point to take into account offender-specific aggravating and mitigating factors. These are factors which do not directly relate to the commission of the particular offence and are generally applicable across all criminal offences; and

  49. (e) Fifth, the court may adjust the individual sentences imposed on the offender to take into account the totality principle.

  50. [37] As explained in Goh Ngak Eng at [45]–[46], these five steps can be classified into two separate stages. Steps one to three allow the court to arrive at an indicative starting point sentence which reflects the “intrinsic seriousness of the offending act”. Pursuant to steps four and five, the court may then adjust the starting point sentence to arrive at a sentence that reflects the “personal circumstances of the offender”.

  51. [38] In proposing the use of the Logachev framework for sentencing in s 5(2) OSA offences, both parties in the present case noted that the Logachev framework has previously been adapted for application in the sentencing of offences other than those punishable under s 172A(2) of the CCA. In Goh Ngak Eng (at [45]–[47]), a three-Judge panel of the General Division of the High Court adapted the Logachev framework for the purposes of sentencing offenders convicted under ss 6(a) and 6(b) of the Prevention of Corruption Act (Cap 241, 1993 Rev Ed) (“PCA”). The court held that the Logachev framework was an appropriate sentencing framework for ss 6(a) and 6(b) PCA offences for two reasons (at [47]). First, the court noted that at the first stage of the Logachev framework, the court would determine the severity of the offending conduct not by reference to the particular facts of each case, but based on the salient features of offending conduct (ie, the offence-specific factors). This made the Logachev framework “apt for encapsulating the diverse circumstances in which private sector corruption can occur”. Second, given the established body of jurisprudence on ss 6(a) and 6(b) PCA offences, the court would be “assisted in identifying the salient features of offending conduct from which the relevant offence-specific factors [could] be derived, and which [would] provide the court with a sense of how the sentencing spectrum under ss 6(a) and 6(b) of the PCA … should be spread across the harm/culpability categories in the sentencing matrix at step two of the framework”.

  52. [39] Implicit in both parties’ submissions in this case was the acknowledgement that the first of the above two reasons applied equally to the sentencing of s 5(2) OSA offences, which – as I have noted – may be committed in a broad variety of circumstances. Further, while the second of the above two reasons might not be applicable in the present case because of the absence of a substantial body of case law on s 5(2) OSA offences, both parties accepted that this was not per se an obstacle to the identification of salient features of offending conduct from which the relevant offence-specific factors for s 5(2) OSA offences could be derived.

  53. [40] I next summarise each side’s submissions as to how the Logachev sentencing framework should be adapted for application to s 5(2) OSA offences.

  54. The Appellant’s submissions on the adaptation of the Logachev sentencing framework for s 5(2) OSA offences

  55. [41] The Appellant proposed that in applying the Logachev sentencing framework to s 5(2) OSA offences, the following offence-specific factors should be considered at the first step of the framework:

  56. [42] As for the second step of identifying the relevant indicative sentencing range, the Appellant suggested that a custodial term should not be imposed for all s 5(2) OSA offences which feature slight harm and low culpability. Per the Appellant’s formulation, the applicable indicative sentencing ranges would be the following:

  57. Harm

  58. Culpability

  59. Low

  60. [43] Further, per the Appellant’s formulation, the number of harm-specific and culpability-specific factors present in each case should determine where in the above matrix that particular case should be located. The Appellant proposed the following categorisation of cases according to the configuration of offence-specific factors present:

  61. (a) 0-1 harm-specific factor and 0-1 culpability-specific factor: slight harm/low culpability category;

  62. (b) 2-3 harm-specific factors and 2-3 culpability-specific factors: categories involving more than slight harm/low culpability to moderate harm/medium culpability (inclusive); and

  63. (c) More than 3 harm-specific factors and more than 3 culpability-specific factors: categories involving more than moderate harm/medium culpability.

  64. [44] Based on the matrix proposed above, the Appellant submitted that the third step of the framework would involve the court identifying an appropriate starting sentence within the selected range, based on – inter alia – the potential prejudice caused to the agency involved.

  65. [45] Next, at the fourth step, the court would consider how (if at all) the sentence derived at the preceding step should be calibrated upon a consideration of offender-specific aggravating and mitigating factors. Any final adjustments needed to account for the operation of the totality principle would then be done at the fifth step. The Appellant also submitted that the offender-specific factors to be considered at the fourth step should include the following:

  66. The Prosecution’s submissions on the adaptation of the Logachev sentencing framework for s 5(2) OSA offences

  67. [46] The Prosecution’s position in respect of the first step of the (adapted) Logachev framework was essentially the same as the DJ’s, in that they adopted the list of offence-specific factors suggested by the DJ at [16] of his GD, with the addition of two other culpability-related factors: firstly, conduct on the part of the receiving party to instigate the communication of the confidential information; and secondly, the receiving party’s persistence in offending.

  68. [47] At the second step of the (adapted) Logachev framework, the Prosecution proposed the following matrix of indicative sentencing ranges:

  69. Harm

  70. Culpability

  71. Low

  72. [48] For the purposes of identifying – at the third step – the appropriate starting point within the relevant indicative sentencing range, the Prosecution submitted that a sentence of a fine would be appropriate only for cases where there were “no aggravating factors” [emphasis in original], having regard to the nature of the information disclosed and the risk of harm. In this context, the Prosecution submitted that offences under s 5 of the OSA had the potential to erode the public’s confidence in the “essential institutions of government”, with the “scale, ease of access and dissemination of information” today only increasing the need for sufficiently robust deterrence.

  73. [49] At the fourth step of the (adapted) Logachev sentencing framework, the Prosecution agreed with the list of offender-specific aggravating and mitigating factors set out at [18] of the DJ’s GD.

  74. [50] In gist, therefore, the Appellant’s and the Prosecution’s positions broadly overlapped. Both parties proposed adapting the Logachev framework so as to establish a five-step sentencing framework for s 5(2) OSA offences. Both parties also agreed that a harm-culpability matrix was an appropriate method by which the severity of the offence could be determined at the first step of the adapted sentencing framework. Parties, however, disagreed on what the indicative sentencing ranges should be at the second step of the adapted sentencing framework; and following from this, they disagreed too on when the custodial threshold would be crossed for s 5(2) OSA offences. There were also minor areas of disagreement regarding the relevant offence-specific factors at the first step of the sentencing framework, and the relevant offender-specific factors at the fourth step.

  75. My decision

  76. [51] Having summarised the parties’ respective submissions, I next explain my decision as to the primary sentencing consideration for s 5(2) OSA offences, the appropriate sentencing framework to be adopted for such offences, and the application of the appropriate sentencing framework to the present case.

  77. General deterrence is the primary sentencing consideration for offences under s 5(2) of the OSA

  78. [52] At the outset, I agreed with the DJ that the primary sentencing consideration for s 5(2) OSA offences is general deterrence. The long title of the OSA states that the OSA is an Act “to prevent the disclosure of official documents and information”. This general legislative purpose is confirmed by the record of the Parliamentary debates regarding the OSA (s 9A(2)(a) read with s 9A(3)(d) of the Interpretation Act 1965 (2020 Rev Ed); see also Tan Cheng Bock v Attorney-General [2017] 2 SLR 850 at [42]–[45]). At [10] of the GD, the DJ referred to the speech by the Minister for Home Affairs, Mr K Shanmugam, which I referenced at [5] above. It is helpful to set out the Minister’s explanation regarding the importance of Singapore’s laws on the handling of official secrets or classified information (Singapore Parliamentary Debates, Official Report (10 January 2011) vol 87):

  79. [53] The above remarks by the Minister make it clear that in Singapore, a laissez-faire response to unauthorised leakages of confidential information will not be tolerated, and that such criminal conduct must be dealt with firmly in order to deter such leakages and to protect the public interest.

  80. Why the Logachev sentencing framework is appropriate for adaptation to s 5(2) OSA offences

  81. [54] Bearing in mind the above sentencing consideration, I address next the formulation of a sentencing framework for s 5(2) OSA offences. I accepted the parties’ submissions that the Logachev framework should be adapted for application to such offences and set out below my reasons. In so doing, I should emphasise that my remarks below are intended to apply only in respect of s 5(2) OSA offences. While it may be said that the offences of unauthorised communication under s 5(1) of the OSA are really the “mirror” of the offences of unauthorised receipt under s 5(2) of the OSA, I did not consider that I should make any pronouncements on the appropriate sentencing approach to s 5(1) OSA offences, given that parties did not make any submissions on the applicability of the same sentencing framework to s 5(1) OSA offences. In any event, it was not necessary for me to make any such pronouncements in dealing with the present appeal.

  82. [55] As I alluded to earlier, offences under s 5(2) of the OSA may be committed in a broad range of circumstances. Section 5(2) of the OSA criminalises the unauthorised receipt of a broad swath of “document[s]” and “information”. The means by which such confidential information may be disclosed and received without authorisation are also manifold; and as the Prosecution pointed out, the corollary of the “modern realities of information systems” is the possibility of vast reams of information being disclosed and disseminated at scale.

  83. [56] From these observations, it follows that in seeking to formulate an appropriate approach to sentencing for s 5(2) OSA offences, it will not be practicable to adopt what is commonly known as the “sentencing benchmark” approach, which proceeds by identifying “an archetypal case (or a series of archetypal cases) and the sentence which should be imposed in respect of such a case” (Terence Ng at [31]). As the Court of Appeal pointed out in Terence Ng (at [32]), the benchmark approach is particularly suited for offences “which overwhelmingly manifest in a particular way or where a particular variant or manner of offending is extremely common and is therefore singled out for special attention”. Section 5(2) OSA offences plainly do not fall within this description.

  84. [57] From the above observations, it also follows that it would not be practicable to adopt what is commonly known as the “sentencing matrix” approach. This is because the “sentencing matrix” approach requires the sentencing court to first determine the seriousness of an offence by reference to a set of principal factual elements of the case in order to give the case a preliminary classification. Based on this preliminary classification, the court then identifies the starting point and the range of sentences before determining the eventual sentence at the second stage of the approach by reference to any other aggravating and/or mitigating factors which do not relate to the principal factual elements (Terence Ng at [33]). The availability of the “sentencing matrix” approach is thus “crucially dependent on the availability of a set of principal facts which can significantly affect the seriousness of an offence in all cases” (Terence Ng at [34]). This has made it appropriate for use in the context of (for example) vice-related offences, where the courts have been able to identify a set of principal factual elements relating to the offender’s role in the vice syndicate (which determines his culpability) and the treatment of the prostitute (which determines the harm caused by the offence): Poh Boon Kiat at [75]–[76]. It is, on the other hand, not suitable for use in relation to s 5(2) OSA offences which can take place in a broad variety of circumstances, making it difficult, if not impossible, to pin-point a set of “principal factual elements” capable of affecting the seriousness of a particular offence.

  85. [58] It is also for this reason that the “single starting point” approach to sentencing would not be appropriate for s 5(2) OSA offences. The “single starting point” approach involves identifying a notional starting point that may then be adjusted based on the aggravating and mitigating factors in each case. This approach is suitable where “the offence in question almost invariably manifests itself in a particular way and the range of sentencing considerations is circumscribed”, as would often be the case for regulatory offences: Terence Ng at [27]–[28]. This does not apply to s 5(2) OSA offences, for the reasons set out in the preceding paragraph.

  86. [59] In similar vein, the “multiple starting points” approach, which involves setting various indicative starting points, each corresponding to a “different class of the offence”, would not be appropriate in this case. One of the usual features of cases to which the “multiple starting points” approach is suited is that a particular mischief is inherent in the offence, which mischief is measurable by reference to a single quantitative metric: Terence Ng at [29]–[30]. This feature is not present in s 5(2) OSA offences.

  87. [60] There is, of course, also the “sentencing bands” approach exemplified in the case of Terence Ng itself. This approach involves a two-stage process of analysis (Terence Ng at [39]; see also Sue Chang v Public Prosecutor [2023] 3 SLR 440 (“Sue Chang”) at [60]). The sentencing court first considers the relevant offence-specific factors to identify the appropriate “band” in which the particular offence should be placed. These offence-specific factors include the manner and mode of commission of the offence as well as the harm caused by it. Having identified the appropriate “band”, the court then considers where within the selected band it would place the offence in question. This first stage of the “sentencing bands” approach results in an indicative starting sentence, which the court proceeds to calibrate at the second stage in consideration of the relevant offender-specific aggravating and/or mitigating factors.

  88. [61] As the High Court noted in Sue Chang at [61], in comparison to the sentencing matrix approach, the sentencing bands approach “is more suitable where the offence can take place in a wide variety of different circumstances, and it is difficult to identify any set of ‘principal factual elements’ which can affect the seriousness of such an offence across the board”. Having considered the elements of the sentencing bands approach, however, I did not find it appropriate for use in the context of offences under s 5(2) of the OSA. The sentencing bands essentially correspond to the number of offence-specific aggravating factors present; and bearing in mind the wide variety of circumstances in which s 5(2) OSA offences may be committed, it is not clear how the sentencing bands approach would accommodate the correspondingly wide range of harm caused by such offences and the potentially significant divergences in the culpability of different offenders.

  89. [62] In contrast, the Logachev framework allows for a more nuanced approach to sentencing in s 5(2) OSA offences. This framework allows the court to classify – at the first step – varying levels of severity of offences “not by reference to the particular facts of each case, but based on the salient features of offending conduct (manifesting themselves as offence-specific factors)”, which allows the “diverse circumstances in which private sector corruption can occur” to be captured: Goh Ngak Eng at [47]. While the court’s comments in Goh Ngak Eng were made in the context of ss 6(a) and 6(b) PCA offences, they are equally apt when applied to offences under s 5(2) of the OSA. For s 5(2) OSA offences, the categories of confidential information and the means of disclosure and receipt of such information are sufficiently wide-ranging to be represented on a harm axis (cf Sue Chang at [70]).

  90. [63] Further, the second and third steps of the Logachev framework allow for a process of “increasing granulation which enhances analytical clarity and promotes the transparent articulation of reasons for the eventual sentence imposed” (Sue Chang at [79]). As the court in Sue Chang observed (at [79]):

  91. It ensures that all relevant sentencing factors are considered and adequate flexibility is built into the approach to allow for sentencing judges to conduct their own assessment and weighing of the relevant harm and culpability factors to arrive at a principled sentence.

  92. [64] I next address the five individual steps of the adapted Logachev sentencing framework for offences under s 5(2) of the OSA.

  93. Relevant offence-specific factors at the first step

  94. [65] The first step of the sentencing framework involves assessing the severity of the offence with reference to the harm caused by the offence and the offender’s culpability. In the table below, I summarise the non-exhaustive offence-specific harm and culpability considerations which are, in my view, relevant to offences under s 5(2) of the OSA:

  95. [66] I first address the harm-specific factors (at [67]–[81] below), followed by the culpability-specific factors (at [82]–[118] below).

  96. (1) Harm-specific factors

  97. (A) Nature of information disclosed (including amount of information disclosed)

  98. [67] Parties agreed that the nature of the information disclosed would constitute a relevant harm-specific consideration, although they disagreed on the precise characterisation of this factor and on some of its constituent elements. At [68]–[74] below, I address five key issues arising from the parties’ submissions.

  99. [68] First, I accepted the Prosecution’s submission that in considering the nature of information disclosed in a particular case, the amount of information disclosed should be taken into account in the sentencing court’s assessment of the seriousness of the offence committed in that case. The Appellant did not dispute this point.

  100. [69] Second, both the Appellant and the Prosecution took the position that the risk of harm caused by the disclosure of the confidential information – while relevant to the court’s assessment of the seriousness of an offence – should not be categorised as an independent harm-specific factor in itself, as it is really a facet of other, existing factors. In my view, this is the correct position. The risk of harm caused by the impugned disclosure can and should be taken into account by the sentencing court in assessing the nature of the information disclosed and the extent of the breach and dissemination – both of which are already distinct harm-specific factors. In considering the nature of the information disclosed, therefore, the court may have to consider (inter alia) the risk of harm arising from the initial disclosure to the immediate recipient of the confidential information (ie, the s 5(2) OSA offender). Then, in considering the extent of the breach and dissemination, the court may also have to consider (inter alia) the risk of harm arising from any further circulation of the information by this s 5(2) OSA offender.

  101. [70] Third, the Appellant suggested that the court should assess the threat posed to the State and third parties by the disclosure of confidential information without authorisation, by taking reference from the classification of information adopted by the Government. In particular, the Appellant referred to Annex A of the Government Data Security Policies (“GDSP”) which provides the following definitions (at p A6):

  102. (a) “Security Classification” is defined as “classifying information according to the security levels that measure the damage done to national security, or national or Agency interest, in the event of an unauthorised disclosure”; and

  103. (b) “Sensitivity Classification” is defined as “classifying information according to sensitivity levels that measure the damage done to an individual or an organisation, in the event of an unauthorised disclosure”.

  104. [71] The Prosecution accepted in principle that any assessment of the nature of the information disclosed should include consideration of the sensitivity of the information, but submitted that reliance on the Government’s classifications should not be taken too far. The Prosecution pointed out that not all information caught by the OSA may be classified; and that importantly, it is in any event the sentencing court which retains the discretion to assess the level of sensitivity of the information disclosed for the purpose of calibrating the seriousness of the offence. In addition, the Prosecution submitted that “sensitivity” – for the purposes of sentencing in s 5 OSA offences – should more accurately be understood as being an inquiry into “whether the information pertains to matters of national importance, including national defence, internal security, public safety, or Singapore’s economic interests”.

  105. [72] Insofar as general principle is concerned, I accepted the Appellant’s submission that the higher the sensitivity classification or security classification of the information leaked, the more likely it is that the risk of harm caused will be higher. However, for the reasons given by the Prosecution (see [71]), the Government’s sensitivity and security classifications should constitute only one sub-factor in the court’s consideration of the nature of information disclosed. This conclusion is reinforced by the qualification found on the first page of the GDSP, which provides that the GDSP “contains general information for the public only”, is “not intended to be relied upon as a comprehensive or definitive guide on each agency’s policies and practices”, and may be updated by the Government “without publishing such updates to the public”. Further, unless the justification for the Government’s classification is admitted in evidence before the sentencing court, the fact per se that the information in question bears a particular classification may not be helpful in practice to the court when it considers the nature of the information disclosed and how it affects the seriousness of the offence.

  106. [73] Fourth, the Prosecution suggested that the “profile of the recipient(s) that the information was disclosed to” is a relevant sub-factor when considering the nature of information disclosed, and accordingly, that the fact that information was disclosed to a “potentially malicious actor” would suggest that the offence committed was more serious. I accepted that the fact that confidential information was disclosed to a potentially malicious actor may be relevant to assessing the risk of harm caused by the said disclosure. It should be highlighted, however, that the profile of the recipient may also be relevant to the court’s consideration of another harm-specific factor – ie, the extent of breach and dissemination. In other words, there may be some overlap in the constituent elements of two distinct harm-specific factors; and the sentencing court will have to exercise care to avoid any double-counting effect.

  107. [74] Fifth, the Appellant suggested that prejudice caused to the “interest of the state” by disclosure and prejudice caused to the “deliberative process” should be subsumed within the analysis of the nature of information disclosed. In my view, however, these matters are more aptly considered within the context of a different harm-specific factor – ie, the extent of actual harm caused. It is this factor which I address next.

  108. (B) Extent of actual harm caused

  109. [76] The Appellant, for her part, did not expressly include the extent of actual harm caused in her proposed list of harm-specific factors. What she did, however, was to include in her list, inter alia, three other suggested harm-specific factors – abuse of information, prejudice to the administration of the State, and the type and extent of harm to third parties. In my view, these three matters really relate to the different types of harm which may be caused by the unauthorised disclosure, and are thus some of the sub-factors to which the sentencing court may have regard when considering the extent of harm caused by the offence in question. In seeking to treat each of these three matters as an independent harm-specific factor, the Appellant’s suggested approach risks introducing over-granularity and excessive prescriptiveness into the sentencing framework (Public Prosecutor v Randy Rosigit [2024] 4 SLR 1586 at [46]). For this reason, I rejected the Appellant’s suggested approach. I preferred the Prosecution’s suggested approach, whereby consideration of the various types of harm is subsumed within the overarching factor of the “extent of actual harm caused”.

  110. [77] I make two other points. I note that in her submissions, the Appellant argued that “[p]rejudice to the agency” – which she equated with the “public service rationale” – should not be considered as a separate harm-specific factor at the first step of the sentencing framework because (according to her), the “public service rationale is inevitably implicated in all offences involving the public sector – OSA offences being such [an] example”. This argument is erroneous. In the first place, “prejudice to the agency” is not at all the same thing as the “public service rationale”. The “public service rationale” is “a sentencing principle which refers to the public interest in preventing a loss of confidence in Singapore’s public administration through corruption within”; and in cases where it is invoked, a custodial sentence is typically justified: Goh Ngak Eng at [66]. Further, depending on the nature of the information and the agency in question, there may be varying types and degrees of prejudice caused to an agency that a sentencing court should consider in assessing the seriousness of the offence. That said, as the Prosecution suggested, such prejudice will form one of the matters weighed in the balance in the sentencing court’s consideration of the extent of actual harm caused, and will not generally be treated as an independent harm-specific factor.

  111. [78] The Appellant also suggested that any public disquiet generated by the offence in question should be relevant to the assessment of harm caused. In the context of an assessment of the harm caused by corruption offences, “public disquiet” has been defined as “the effect of causing loss to third parties” or “generating a sense of unease in the general public”: Wong Chee Meng at [67], cited in Goh Ngak Eng at [60(c)]. Applying a similar definition in the present case, I was satisfied that the various types of harm identified by the Prosecution (see [75] above) would be sufficiently broad to include consideration of any public disquiet caused by the unauthorised receipt of confidential information.

  112. (C) Extent of breach and dissemination

  113. [79] The third harm-specific factor I have set out is the extent of breach and dissemination. Parties agreed that the number of persons to whom the information was leaked without authorisation (including the number of third-party recipients) would be a relevant harm-specific factor. As alluded to earlier (at [69]), the risk of harm caused by any subsequent dissemination beyond the initial recipient of the confidential information (ie, the s 5(2) OSA offender) may be considered within the context of the extent of breach and dissemination.

  114. [80] In this connection, I accepted the Prosecution’s submission that the relevant sub-factors in relation to the extent of breach and dissemination are the medium and scale of dissemination. All other things being equal, greater harm is likely to be actually caused if the disclosure of official information without authorisation takes place via publicly accessible media (as opposed to private channels), and/or if such disclosure results in dissemination to a large group of persons.

  115. [81] I also accepted the Appellant’s submission that “[e]ven though the secondary spread of protected information by third parties may be beyond an offender’s control, it is not precluded from being relevant to the determination of punishment for that offender” (citing Lai Jenn Wuu v Public Prosecutor [2013] 4 SLR 1134 at [4]). In other words, the degree of further dissemination by third parties is a relevant sub-factor in the court’s assessment of the extent of breach and dissemination.

  116. (2) Culpability-specific factors

  117. [82] I next address the offence-specific factors that are relevant to an offender’s culpability.

  118. (A) Offender’s role

  119. [83] An offender’s role in the commission of the offence for which he is being sentenced has generally been held to be an important indicator of his culpability: Logachev at [61]. For example, in cases involving a criminal syndicate, the higher the offender’s position in the syndicate, the more culpable he is likely to be (Logachev at [60]). As another example, an offender who actively takes steps to solicit the unauthorised disclosure of confidential information will generally be found to be more culpable than a passive recipient who took no steps to pursue the acquisition of the information (Goh Ngak Eng at [84]).

  120. [84] In this connection, I note that in the Prosecution’s written submissions, it was suggested that the court should consider, as an independent culpability-specific factor, whether an offender had initiated the unauthorised communication of information. In my view, this is not the correct approach. In the context of corruption offences, it has been observed that “it is preferable to examine holistically whether the offender’s role in the corrupt transaction had been active or passive, rather than to focus restrictively on whether he had initiated or solicited the corrupt scheme or the bribe” (Goh Ngak Eng at [84]). This reasoning applies with equal force to offences under s 5 of the OSA. The fact that a recipient of information initiated its unauthorised communication is relevant to assessing his culpability only insofar as it demonstrates the degree to which he played an active role in the entire illicit transaction.

  121. [85] I also note that in its submissions, the Prosecution employed the terms “instigate” and “initiate” interchangeably, apparently on the basis that they meant the same thing. Again, this is not correct. The term “instigate” is given a specific meaning in criminal law: see Explanation 1 to s 107 of the Penal Code 1871 (2020 Rev Ed) (“Penal Code”). The Prosecution plainly did not intend to invoke the specific meaning given to this term in the Penal Code; I address this further at [150] below. As such, using this term interchangeably with the term “initiate” in a loose or colloquial manner may simply create unnecessary confusion.

  122. [86] I make one final point in relation to this culpability-specific factor. In its submissions, the Prosecution accepted the DJ’s finding that the “position held by the offender” formed a relevant consideration alongside the offender’s role. More precisely, the Prosecution submitted that a sentencing court should consider – as distinct culpability-related factors – the “[p]osition held by the offender, such as whether he is a senior public officer, or in a sensitive vocation such as military intelligence”, as well any “[a]buse of trust, including the degree to which the offender exploited their position of trust, access privileges or security clearance”. In my view, however, considerations of “abuse of position” and “breach of trust” are more appropriately raised in the context of sentencing in s 5(1) OSA offences, where the offender is usually someone in an official position that allows him access to the confidential information in question. In the context of s 5(2) OSA offences, there may, of course, be offenders who seek to take advantage of their relationship with a person in an official position in order to obtain such information without authorisation. Such conduct can and should be taken into account by the court when it considers the existing factor of the s 5(2) OSA offender’s role. I add that such conduct may also be relevant for consideration in the context of the separate culpability-specific factors of planning and premeditation as well as the level of sophistication: the sentencing court will need to exercise care to avoid double-counting.

  123. (B) Duration and persistence of offending behaviour

  124. [87] In respect of the duration and persistence of the offending behaviour, I accepted the Appellant’s and the Prosecution’s submission that this constitutes a relevant culpability-specific factor. An offence that is perpetrated over a sustained period, or on multiple occasions, will generally be more aggravated than a one-off offence: Logachev at [59], affirmed in Public Prosecutor v GED [2023] 3 SLR 1221 at [78].

  125. [88] In this connection, I note that the parties were divided on the issue of whether events occurring after the offence in question could be taken into account by the sentencing court in its assessment of the offender’s culpability. In the proceedings below, the DJ justified the imposition of a custodial sentence on the basis (inter alia) of the incident in September 2022 in which the Appellant had asked the co-accused whether a “shopfront” she intended to rent “had previously been ‘checked’ by the Police”. The DJ held that this incident amounted to a “subsequent attempt to commit the same offence some two years later”, and that although no separate charge had been preferred in relation to this incident, it “probably constituted an offence under s 511 of the [Penal Code]”. In his view, this incident demonstrated that the s 5(2) OSA offence for which he was sentencing the Appellant “was not a momentary lapse of judgment”.

  126. [89] On appeal, the Appellant argued that the DJ erred in principle in finding the September 2022 incident to be a relevant consideration in his assessment of her culpability. The Prosecution, on the other hand, submitted that Public Prosecutor v Bong Sim Swan Suzanna [2020] 2 SLR 1001 (“Suzanna Bong”) provided authority for the DJ’s decision to consider subsequent uncharged offending conduct when assessing the Appellant’s culpability. In response to the Appellant’s argument that the decision in Suzanna Bong dealt solely with “prior uncharged offending conduct”, the Prosecution contended that the principles espoused in that case were just as applicable to conduct subsequent to the offence for which an offender had been charged.

  127. [90] I found the Prosecution’s position to be only partially correct. At the outset, it must be recognised that as a matter of general principle, an offender cannot be punished for an offence for which he was not charged or convicted. Indeed, in the context of prior uncharged offending conduct, the Court of Appeal in Suzanna Bong took pains to highlight (at [64]) that “[i]f the Prosecution wants the sentencing court to consider the offender’s prior acts and these acts might also be the subject of criminal charges, the onus is on them to draw up the necessary charges and proceed with them at trial or apply for them to be taken into consideration for the purpose of sentencing pursuant to s 148 of the CPC”.

  128. [91] There are at least two main reasons for this general principle. As the High Court in Public Prosecutor v Tan Thian Earn [2016] 3 SLR 269 (“Tan Thian Earn”) observed (at [62]), not punishing an offender for an offence for which he was not charged “is an elementary component of fairness”. In addition, there is “a constitutional dimension to this issue”: the decision whether to frame a charge and if so, what charge to frame, is the constitutional prerogative of the Public Prosecutor; and in a scenario where (for example) the Public Prosecutor chooses not to frame a charge for each of the prior acts of offending, the court should not be asked to “indirectly sanction” an offender for those prior acts by way of an enhancement to the sentence for the offence in respect of which a charge was framed. Instead, if the Prosecution desires a prior offending act to be taken into consideration by the court in sentencing the offender, it should draw up an appropriate charge in respect of that prior act. If the Prosecution elects not to do so or is unable to draw up an appropriate charge (eg, due to lack of evidence), there is no reason why it should expect to be entitled to ask for the prior act to be taken into account in sentencing.

  129. [92] In Chua Siew Peng v Public Prosecutor [2017] 4 SLR 1247 (“Chua Siew Peng”), the High Court elaborated on the above concerns. In considering whether an offender’s past offending conduct could be considered for the purposes of sentencing, the court noted that if the Prosecution drew up charges in respect of an accused’s past offending conduct but then elected not to proceed with these charges, it must apply for these outstanding charges to be taken into consideration (“TIC”) if it wishes the court to consider them in sentencing the accused for the charges proceeded with. Pointing out that the CPC had provided “legal safeguards” for how TIC offences could be considered at the sentencing stage, the court expressed the concern that if there were outstanding charges which were not TIC but which were nevertheless considered by the court at the sentencing stage, this would in effect amount to the court “recognising a new category of offences, to be taken into consideration for the purpose of sentencing at the sentencing stage, in addition to the legislatively provided category of TIC offences under s 148 of the CPC”. In the court’s view, this would be “an undesirable outcome”: Chua Siew Peng at [78].

  130. [93] While the above concerns were articulated by the High Court in Chua Siew Peng (at [78]) in the context of past offending conduct for which the Prosecution had drawn up the necessary charge which it then elected not to proceed with or to apply to TIC, similar concerns are a fortiori relevant in the context of offending conduct occurring at a different time from the charged offence (whether before or after it) for which the Prosecution has not drawn up any charges.

  131. [94] At the same time, while bearing in mind the importance of the above principle, our courts have acknowledged that a sentencing court should consider all the circumstances of a case when it assesses the seriousness of the offence in question. In this connection, as the Court of Appeal explained in Suzanna Bong at [73], “[w]here the Prosecution has proved relevant facts”, there is no reason “why the court should pay no heed to them when considering the appropriate sentence on the sole ground that they might also amount to offences”. As the Court of Appeal put it (at [73]):

  132. … [I]t is important to consider the totality of the circumstances of a charged offence in order to have a true flavour of the offence as the overall perspective may have an impact on the level of the offender’s culpability and the extent of the victim’s suffering. Naturally, in applying this principle, the court must take a common-sense and contextual approach when considering the importance of the proved relevant facts.

  133. [95] As to how the sentencing court is to consider “the totality of the circumstances of a charged offence” without making unacceptable inroads into the general principle that an offender cannot be punished for an uncharged offence, the preceding passages at [65]–[66] of the judgment of the Court of Appeal in Suzanna Bong are instructive. In these passages, the Court of Appeal cited with approval the decision of the High Court in Chua Siew Peng:

  134. [65] In Chua Siew Peng … the High Court observed that while a sentencing court generally could not take into account uncharged offences, it was entitled to and in fact should consider the aggravating circumstances in which the offence was committed, even where those circumstances could technically constitute separate offences (Chua Siew Peng at [81]). There was conduct that could constitute a separate offence but which was so closely intertwined with the specific charge before the court that it should be considered at sentencing (Chua Siew Peng at [83]). One example was the offence of drink-driving where the sentencing court might recognise aggravating factors such as speeding or driving recklessly, notwithstanding that each of those facts could amount to a separate charge (Chua Siew Peng at [83]). A fact with a sufficient nexus to the commission of the offence could be considered at the sentencing stage, irrespective of whether this fact could also constitute a separate offence for which the accused was not charged. What constituted a sufficient nexus was a fact-sensitive inquiry that depended on the circumstances of each case and the degree of proximity of time and space to the charged offence. A sufficient nexus would generally be present if it concerned a fact in the immediate circumstances of the charged offence or was a fact relevant to the accused’s state of mind at the time of committing the offence (Chua Siew Peng at [85]).

  135. [66] We agree with the above principles stated in Chua Siew Peng. If the facts are relevant and proved, they may be, and indeed ought to be, considered by the sentencing court …

  136. [96] Having endorsed the principles stated by the High Court in Chua Siew Peng, the Court of Appeal in Suzanna Bong went on to consider how these principles had been applied in that case. The offender in that case (Chua) was convicted of voluntarily causing hurt to and wrongful confinement of a domestic worker in her employment. She was charged with slapping the victim repeatedly on 29 October 2012. During the incident, she had also pulled the victim’s hair. The next day, she locked the victim in the house; and there was evidence that prior to this occasion, she had locked the victim in the house on many other occasions. On this occasion, the victim climbed out of the window to escape, sustaining multiple fractures in the process. The High Court found that Chua’s act of pulling the victim’s hair – although not part of the charge of voluntarily causing hurt – was an aggravating factor “because it occurred contemporaneously with the slap and formed part of the immediate circumstances of the offence” (Suzanna Bong at [69], citing Chua Siew Peng at [67] and [87]). As for the previous instances of wrongful confinement, these were relevant to the court’s consideration of Chua’s culpability because “[i]n knowingly prolonging the victim’s wrongful confinement, [Chua] increased the risk that the victim would suffer injuries owing to the conditions of the confinement”, and her actions ultimately “drove the victim to take the drastic step of escaping through a window”: Suzanna Bong at [69], citing Chua Siew Peng at [65]. As the High Court in Chua Siew Peng pointed out at [91], the past instances of confinement had “a close nexus” to the offence of wrongful confinement for which Chua was charged, insofar as “they had made the [v]ictim’s mental state more vulnerable”: the offence in question was thus committed knowingly by Chua on a victim with a significantly weakened mental state.

  137. [97] Following from the reasoning of the Court of Appeal in Suzanna Bong and its endorsement of the principles set out in Chua Siew Peng, conduct amounting to an uncharged offence may be taken into account by a sentencing court if such conduct is shown to have a sufficient nexus to the charged offence. Further, while the question as to what may amount to a sufficient nexus is a fact-sensitive inquiry, as the High Court noted in Chua Siew Peng (at [84]), the degree of proximity of time and space between the uncharged offending conduct and the charged offence will be a pertinent factor in this inquiry. The greater the temporal and/or spatial distance between the uncharged conduct and the charged offence, the less likely it is that the uncharged conduct will have “sufficient nexus” to the commission of the offence.

  138. [98] GDC v Public Prosecutor [2020] 5 SLR 1130 (“GDC”) provides an example of this fact-sensitive inquiry. In that case, the accused was originally charged with and convicted by the District Judge of aggravated outrage of modesty of a person under 14 years of age (which was an offence punishable under s 354A(2)(b) of the Penal Code (Cap 224, 2008 Rev Ed)). On appeal, Sundaresh Menon CJ amended this charge to one of outrage of modesty of a person under 14 years of age, under s 354(1) read with s 354(2) of the Penal Code (Cap 224, 2008 Rev Ed) and convicted the accused of the amended charge. A crucial aspect of this case was that shortly after outraging the modesty of the victim, the accused had slapped the victim’s face twice: GDC at [5]. Although this act of slapping the victim did not form part of the amended charge in GDC, it was taken into account as an offence-specific aggravating factor despite it having occurred after the commission of the offence of outrage of modesty. In reaching this conclusion, Menon CJ noted that the slapping of the victim “took place just ten minutes later in the context of the same struggle [during the course of which the accused outraged the victim’s modesty] in the same place” (GDC at [43], citing Chua Siew Peng and Suzanna Bong). The decision in GDC thus reinforces the proposition that uncharged post-offence wrongful conduct will generally be found to have sufficient nexus to the charged offence where such uncharged conduct is relatively proximate in time and space to the charged offence.

  139. [99] For completeness, I note that in Public Prosecutor v Muhammad Salihin bin Ismail [2023] SGHC 155 (“Salihin (HC)”) at [109], the High Court held that conduct (in this case, the accused’s delay in seeking medical attention for the victim) which occurred after the commission of the offence (voluntarily causing grievous hurt, “VCGH”) should not be taken into account as an aggravating factor. The Court of Appeal subsequently allowed the Prosecution’s appeal against the High Court’s decision to acquit the accused of the original murder charge and to convict him of a lesser charge of VCGH. In convicting the accused of the original murder charge and sentencing him to life imprisonment and caning, the Court of Appeal did not comment on the High Court’s treatment of the accused’s post-offence conduct in sentencing (see Public Prosecutor v Muhammad Salihin bin Ismail [2024] 1 SLR 792). Having perused the High Court’s judgment in Salihin (HC), I was of the view that its treatment of the accused’s post-offence conduct was not necessarily in conflict with my observations at [90]–[98] above. While the High Court did express the view that the accused’s delay in seeking medical treatment for the victim should not be treated as an aggravating factor for sentencing purposes (at [109]), it went on to consider (at [110]) what the outcome should be if such delay were to be considered an aggravating factor: the court’s decision was that this factor was – in any event – “not engaged on the facts of the case” because there was “insufficient evidence” to support the finding that the delay had been deliberate. Additionally, from the judgment (at [109]), it is apparent that in deciding not to treat the accused’s delay in seeking medical treatment for the victim as an aggravating factor for sentencing purposes, the court was influenced – at least in part – by its finding that:

  140. …to the extent that a delay in seeking medical attention contributed to the victim’s demise or the more severe injury that he or she eventually suffered from, this already constituted a consequence of the offence, which would have been factored in identifying the indicative starting sentence under the first step of the sentencing framework [for VCGH offences] in [Public Prosecutor v BDB [2018] 1 SLR 127 (“BDB”)] (see BDB at [56]).

  141. [100] To sum up, therefore, in considering the duration and persistence of offending behaviour as a culpability-specific factor in s 5(2) OSA offences, sentencing courts may consider facts with sufficient nexus to the commission of the charged offence even if such facts occurred after the commission of the charged offence and even if they are not the subject of a separate charge.

  142. (C) Extent of offender’s knowledge

The judgment continues past this page. Read the rest at eLitigation.