Who’s watching Mr Big? Scenario operations and induced confessions

ABSTRACT

Introduced in Australia in the late 1990s, the Mr Big investigative technique is a complex undercover police operation structured around a fictional criminal gang, which is created to build rapport with the suspect of a serious crime. The suspect is encouraged to confess to the crime to benefit from advantages and protection offered by the gang, including the possibility of influencing the outcome of any prosecution. Challenging Mr Big confessions in Australia has been unsuccessful. The lack of transparency and need for deception inherent in covert operations, however, exposes a potential for fabricated and coerced confessions and the risk of wrongful convictions. Scenario operations are not directly scrutinised by independent oversight bodies and an examination of reported cases suggests that Australian Mr Big operations do not involve covert police in unlawful activities that might been scrutinised under controlled operations regimes. This is a concern given a cultural predisposition within Australian law enforcement towards secrecy, rank-closing, and resistance to examination. We explore options for ensuring the safety of Mr Big convictions, including implementation of robust, independent oversight mechanisms, amendments to evidentiary rules about inducement-based confessions, and revised warning regimes to include direction on the need for caution in accepting scenario confessions.

KEYWORDS:

• covert operations
• false confessions
• Mr Big confession
• Mr Big operations
• scenario evidence

Introduction

The covert police investigative technique known as a ‘Mr Big scenario operation’ has attracted scrutiny over concerns about the reliability of confessions obtained through its use (Luther & Snook, 2016, p. 18; Murphy & Anderson, 2016). Originating in Canada in the 1990s, Mr Big operations are typically used as a last resort where there is insufficient evidence to charge a suspect (Dufraimont, 2015, p. 477). The tactic involves undercover police posing as members of a fake criminal gang with the aim of eliciting a confession from a suspect as part of their recruitment into the group (Connors et al., 2017; Murphy & Anderson, 2016).

Mr Big operations are different to other undercover operations, which usually involve police infiltrating existing criminal organisations to obtain evidence of criminal dealings (Iftene, 2019, p. 31). In a Mr Big operation, the criminal organisation is created by police for the sole purpose of obtaining a confession (Glazebrook, 2015). Covert police officers assume all roles in the fake criminal gang, including the role of the organisation boss, Mr Big (MacLean & Chapman, 2016, p. 1). Mr Big operations typically begin with an undercover officer befriending the suspect and offering them easy work for reward. Over several weeks, the suspect becomes increasingly involved with the organisation through a series of elaborate scenarios, which also reveal the organisation’s apparent power and wealth (Luther & Snook, 2016, p. 132; Phelan, 2019).

The operation culminates in a meeting with Mr Big who uses persuasive reasoning to press the suspect to confess to the crimes under investigation (Iftene & Kinnear, 2020, p. 302; Palmer, 2005, p. 112). The suspect will be told they must satisfy Mr Big that they would not bring unwanted police attention to the organisation, and that the organisation can make problems such as police investigations ‘go away’ through its access to (fictional) corrupt police officers (Tofilau v The Queen (2007) 231 CLR 396, 410; Lauchlan v The State of Western Australia [2008] WASCA 227, [17]). The suspect may also be under the impression that they will lose the organisation’s camaraderie and access to wealth if they do not confess (e.g., R v Marks [2004] VSC 476; Lauchlan v The State of Western Australia [2008] WASCA 227). On confessing, the target is arrested and formally interviewed.

Covert methods of policing are regarded as a necessary function of policing (Loftus, 2019; Ross, 2007) and sometimes involve conduct that would normally be illegal (Bronitt, 2004; Grabosky & Urbas, 2019). In Australia, the potential for abuse is ostensibly mitigated by oversight bodies, controlled operations legislation authorising illicit actions by police, and judicial gatekeeping regarding the admissibility of evidence. The adequacy of these oversight mechanisms is questionable, however, given many (if not most) Mr Big operations in Australia are not ‘controlled operations’, and key evidentiary rules about admissions do not apply. Mr Big targets have been unsuccessful in having confessional evidence excluded at trial because courts have consistently held that rules prohibiting inducement-based confessions do not apply to covert police. This approach has followed Tofilau v The Queen (2007) 231 CLR 396 in which the High Court upheld the admissibility of confessions elicited through the Mr Big technique.

While there are no established wrongful convictions in Australia or Canada resulting directly from Mr Big operations (Connors et al., 2019, p. 3; Moore et al., 2009), there are longstanding concerns about the potential for false confessions, which are a known risk factor in wrongful convictions (Iftene & Kinnear, 2020; Lackey, 2020; R v Hart [2014] 2 SCR 544, [62] (Justice Moldaver)). For instance, Mr Big suspects are not afforded fundamental rights available to other suspects, such as the right to remain silent, because the meeting with Mr Big is not a formal police interview (Glazebrook, 2015; Iftene, 2019; R v Hart [2014] 2 SCR 544, [64]). In a dissenting judgement in Tofilau, Justice Kirby observed that the Mr Big technique allowed undercover police to ‘circumvent’ the need to uphold an accused’s right to silence, and right to be warned about the consequences of self-incrimination (Tofilau v The Queen (2007) 231 CLR 396, 443). Justice Kirby pointed to the tendency for scenario operations to select vulnerable targets (444). This concern was echoed by the Supreme Court of Canada in the landmark case of R v Hart [2014] 2 SCR 544, which noted the potential for Mr Big operations to become coercive and prey on suspects’ vulnerabilities, threatening the integrity of the criminal justice system ([117]). Recent research has also found a relationship between higher levels of suggestibility and compliance in Mr Big suspects and an increased susceptibility to making false confessions (Otgaar et al., 2021).

The Supreme Court of Canada acknowledged the potential unreliability and prejudicial effect of Mr Big confessions and introduced a new two-pronged approach to determine their admissibility (R v Hart [2014] 2 SCR 544, [10]). Under this approach, Mr Big confessions are presumptively inadmissible unless the probative value of the confession outweighs its prejudicial effect on the balance of probabilities. The second prong shifts the onus to the defence to establish an abuse of process (R v Hart [2014] 2 SCR 544, [89]).

Australian courts have distinguished Australian Mr Big operations from the features of Canadian operations that resulted in the Hart judgement. In R v Jelicic [2016] SASC 57, Justice Peek observed that unlike the Canadian operations, the ‘Australian methodology’ required all contact between the suspect and the covert police to be aurally recorded ([123]). The Canadian methodology also often involved covert police deliberately creating an impression of the organisation’s willingness to use violence, whereas Australian operations have emphasised to suspects that their cooperation was voluntary, and they could leave freely (Deakin v The Queen (2019) NTCCA 21, [76]; see also R v Rumsby [2023] NSWSC 229, [43] and R v Cowan [2015] QCA 87, [85]–[87]).

The 2020 Report of the Royal Commission into the Management of Police Informants in Victoria, however, has highlighted difficulties in balancing police independence in their investigative roles with appropriate scrutiny to limit opportunities for abuse (RCMPI, 2020). This balancing act is particularly problematic in covert police operations involving deception or the use of informants (Grabosky & Urbas, 2019; Ross, 2007).

This article presents findings from our study of 22 reported Mr Big cases. We begin by considering issues in the oversight of Mr Big Operations. We explore the implications of excluding covert police from prohibitions against inducement-based confessions, the desirability of seeking supportive evidence of the matters confessed to and explore safeguard options. We argue that greater scrutiny of Mr Big confessions is needed in Australia in the wake of the Royal Commission revealing Victoria Police’s ‘win at all costs’ approach and significant non-disclosure practices.

Finding Mr Big

Introduced in Australia by Victoria Police in the late 1990s, Mr Big tactics were initially prioritised for old cold case investigations but eventually used in numerous Victorian cases including R v Favata [2004] VSC 7; R v Tofilau [2003] VSC 188; R v Hill [2004] VSC 293; and R v Marks [2004] VSC 476. With the assistance of Victoria Police, the technique has been employed across Australia.

We identified Mr Big cases for this study by searching the AustLII and Westlaw AU legal databases. Legal data base searching, however, can only provide a limited understanding of the scope of Mr Big operations as not every operation will have been successful in securing a confession.

Cases were initially located using search terms gleaned from the literature: ‘Mr Big’, ‘covert operation’, ‘scenario method’, ‘scenarios’ ‘scenario technique’, ‘covert police’, ‘scenario evidence’, and ‘Tofilau v The Queen’. We applied additional search terms obtained from the different terminology used by Australian courts to refer to Mr Big operations: ‘covert police operation’ (R v Cowan (2013) 237 A Crim R) ‘scenario investigation’ or ‘Mr Big investigatory strategy’ (R v Kilincer (No. 2) [2021] NSWSC 829); ‘Mr Big’ or ‘Big Boss’ operation; the ‘Canadian technique’ (R v Simmons (No. 5) [2015] NSWSC 333); ‘undercover operation of the type considered in Tofilau v The Queen’ (Donai v The Queen [2011] NSWCCA 9); ‘homicide scenario technique’ (Lauchlan v The State of Western Australia [2008] WASCA 227); ‘Mr Big covert operation’ (R v Jelicic [2016] SASC 57); and ‘Undercover Serious Crime Undercover Technique’ or ‘USCUT’ (R v Rumsby [2023] NSWSC 229). We searched for cases at multiple instances and excluded cases that did not involve a confession or a reference to a covert police operation involving the use of scenarios, or the creation of a fake criminal gang. This process resulted in twenty-two cases, which are represented in Table 1.

Table 1. Mr Big cases.

Jurisdiction	Case	Verdict
VIC	DPP v Ghiller (2003) 151 A Crim R 148	Convicted of armed robberies and other offences
	R v Clarke [2004] VSC 11	Convicted of murder
	R v Tofilau (2003) 13 VR 1; Tofilau v The Queen (2007) 231 CLR 396	Convicted of murder
	R v Favata [2004] VSC 7	Convicted of murder
	R v Hill [2006] VSCA 41	Convicted of murder
	R v Marks (2004) 150 A Crim R 212	Convicted of murder
	DPP v Bennett [2004] VSC 207	Convicted of murder
	R v Karakas (Ruling No. 1) [2009] VSC 480	Convicted of murder
	R v Weaven (Ruling No. 1) [2011] VSC 442	Convicted of murder
	Kamalasanan v Queen [2019] VSCA 180	Convicted of murder
WA	The State of Western Australia v Lauchlan [2005] WASC 266; Lauchlan v The State of Western Australia [2008] WASCA 227	Convicted of murder
NSW	Donai v The Queen [2011] NSWCCA 9*	Convicted of murder
	R v Abrahams [2013] NSWSC 952*	Convicted of murder and improperly interfering with a body
	R v Simmons; R v Moore (No. 2) [2015] NSWSC 143	Simmons was found not guilty of murder but guilty of assault occasioning actual bodily harm (R v Simmons (No. 7) [2015] NSWSC 574)
	R v Fesus (No. 2) [2015] NSWSC 1467	Convicted of murder
	R v Kilincer (No. 2) [2021] NSWSC 829	Convicted of murder
	R v Rumsby [2023] NSWSC 229	Convicted of murder
TAS	Standage v Tasmania [2012] TASSC 88; [2017] TASCCA 23	Convicted of murder
QLD	R v Cowan (2013) 237 A Crim R 388; [2016] 1 Qd R 433	Convicted of murder
	R v Taylor [2016] QSC 116	Convicted of murder
SA	R v Jelicic [2016] SASC 57	Convicted of murder
NT	Deacon v The Queen (2019) 282 A Crim R 303	Convicted of murder

*Referred to in R v Simmons (No. 5) [2015] NSWSC 333.

It is impossible to determine how many operations have been conducted across Australia because, as we discuss in the next section, police do not publicly reveal the details of their operations and oversight bodies are precluded from disclosing police methodologies (see for example, Law Enforcement (Controlled Operations) Act 1997 NSW, s 24(1)(b)). Koningisor described police secrecy protection as a kind of ‘exceptionalism’ in that police are treated differently than other executive government agencies when it comes to transparency obligations, comparable to the exceptional protections afforded national security secrecy claims (Koningisor, 2023, p. 637, 662).

Oversight regimes

Australian Mr Big operations have been assumed to be ‘controlled operations’ (e.g., Murphy & Anderson, 2016; Phelan, 2019) but there is no publicly available information about which operations involved controlled activities, if they do at all. Controlled operations are a well-established law enforcement methodology in Australia and involve police deliberately enabling and participating in a crime (Joint Committee on the National Crime Authority, 1999; Murphy, 2014, p. 38). A controlled activity is an activity that would be unlawful unless authorised by an appropriate authority.

Until 1995, police relied on the exercise of judicial discretion to admit evidence obtained through covert operations. In Ridgeway v The Queen (1995) 184 CLR 19, Australian Federal Police and Malaysian police organised a controlled importation of heroin into Australia to apprehend Ridgeway. The importation was an offence under Section 233B (1) of the Customs Act 1904 (Cth) and the defence challenged the reception of the evidence. The High Court determined that the importation was illegal and evidence of it should have been excluded on public policy grounds. The controlled importation might be an acceptable law enforcement technique, but law enforcement officers could only legitimately engage in this kind of conduct if authorised under legislation (Ridgeway v The Queen (1995) 184 CLR 19, 43–44 per the majority; 53–54 (Brennan J)).

The Ridgeway decision triggered the introduction of legislation regulating controlled operations (e.g., Law Enforcement (Controlled Operations) Act 1997 (NSW)) and oversight regimes across Australia. These oversight regimes disclose minimal information about controlled operations. For instance, the NSW Office of the Inspector of the Law Enforcement Conduct Commission (OILECC) is required to inspect records of NSW Police Force to determine whether the requirements of the Act are being complied with and report to Parliament (OILECC, 2022, p. 4). The reports include the number of controlled authorities granted or varied, the number of participants, the nature of the suspected criminal activity and any controlled activities (e.g., supplying prohibited drugs). There is no information about the type of operation. In the 12-month period ending 30 June 2022, the OILECC simply reported that 268 controlled operations authorities were granted involving 4649 law enforcement participants, of whom 303 were engaged in controlled activities (pp. 35–36).

As indicated, Australian Mr Big operations are not necessarily controlled operations. The operations described in the cases in this study were simulated or ‘staged’ criminal activities rather than controlled activities (e.g., DPP v Ghiller [2003] VSC 350, [64]; R v Rumsby [2023] NSWSC 229, [50]). The courts in R v Taylor [2016] QSC 116, [19] and R v Kilincer (No. 2) [2021] NSWSC 829, [22] made a point of stating that police did not seek authorisation to run controlled operations in these cases because there was never an intention that the scenarios involve criminality.

It is also possible that no Australian Mr Big operation has ever been authorised as a controlled operation. As discussed above, the Australian methodology differs from Canadian Mr Big Operations, which had used implied threats of violence towards the suspect as a tactic to secure a confession (R v Cowan [2015] QCA 87, [85]–[87]). An early decision to rely on simulated criminal activities and emphasise the voluntary involvement of suspects may have resulted from the need for caution in the wake of Ridgeway. When the methodology was introduced to Australia in 1999 by Victoria Police, only the Commonwealth, New South Wales and South Australia had enacted legislative controlled operations authorisation regimes (Joint Committee on the National Crime Authority, 1999).

Covert operations not involving controlled activities are authorised by police command or committee. Police annual reports, however, do not disclose any details about Mr Big operations. While Australian jurisdictions have reporting requirements for the assumed identity and covert surveillance authorisations that usually accompany Mr Big operations, once again these provide scant information. For instance, the Commissioner for the NSW Crime Commission is required to provide annual reports on authorities for assumed identities including a ‘general description of the activities undertaken by authorised persons when using assumed identities under the Act’ (Law Enforcement and National Security (Assumed Identities) Act 2010, s 35(1)(a)). The 2021–2022 NSW Crime Commission Annual Report simply disclosed that one authority was granted for the period.

Mr Big operations require comprehensive oversight and accountability to protect against false confessions (Puddister & Riddell, 2012). Oversight of covert police investigations in Australia, however, is patchwork. The OILECC recently reviewed inconsistencies in NSW’s legal framework relating to the use and oversight of covert investigative powers, including the lack of consistency in external oversight regimes. The report noted that, of the suite of covert powers available to law enforcement, only decisions to grant authorisation for controlled operations and assumed identities were made internally (OILECC, 2022, p. 49). Authorisations for other covert operations such as searches, surveillance, device, and telephone intercepts were made externally under warrant and scrutinised externally. The OILECC recommended a comprehensive review of the legal framework (2022, p. 65). Earlier in a submission to the Royal Commission into the Management of Police Informants in Victoria, the Independent Broad-based Anti-corruption Commission (IBAC) cautioned that the current oversight system resulted in a ‘lack of transparency and consistency, alongside increased risks of monitoring [gaps] that undermine public confidence in the oversight of covert and intrusive powers of Victoria Police’ (RCMPI, 2020, p. 210).

This lack of transparency includes a dearth of information about the cost of Mr Big operations and ‘value for money’ considerations. Mr Big operations are expensive, involving large numbers of police operatives over extended periods of time. The investigation into Brett Cowan involved 36 covert operatives (R v Cowan [2015] QCA 87, [29]). The costs of running these complex operations include accumulated travel, accommodation, food expenses and set up costs as well as any payments and benefits made directly to the target (Luther & Snook, 2016). Deacon’s receipt of $4700 in cash payments over 33 scenarios as well as ‘travel and other benefits’ is indicative of the typical outlay to a target (Deacon v The Queen [2019] NTCCA 21, [12]). Canadian operations are reported to have costed as much as $4 million CAD (Glazebrook, 2015).

While some information about money paid to the operation targets can be gleaned from published cases, information about the costs of running a Mr Big Operation in Australia is not publicly available. These costs are hidden within financial reporting by police such as the 2021–2022 Annual Report of the Queensland Police Service, and potentially within financial reporting of controlled operations. Only Queensland publishes financial information about controlled operations through reports by the Queensland Controlled Operations Committee’s (QCCC) independent member who has inspection and reporting functions in relation to controlled operation records maintained by the Queensland Police Service (Police Powers and Responsibilities Act 2000 (QLD) ss 269, 272). In the reporting period of 01 July 2021–30 June 2022, eight new operations were approved and five finalised with a total operational expenditure of $168,846.08 (QCCC, 2022). The five finalised operations resulted in the arrest of 38 individuals and the laying of 272 charges but there is no information about the individual costs or nature of each operation (QCCC, 2022, p. 3).

Harfield suggested that an effective governance framework for accountable policing can be achieved through good internal governance, external scrutiny, and the rules of evidence (2010, p. 776). As indicated above however, the adequacy of internal governance of Australian Mr Big operations is unknown as this information is not revealed in annual reporting requirements and police methodologies are generally protected from disclosure. Further, the audit function provided by current oversight regimes is not concerned with the conduct of Mr Big operations or the use of deceptive methodologies but rather ‘the investigator-generated records of that conduct that are audited in lieu of direct governance’ (Harfield, 2010, p. 779). This leaves the courts providing the only external oversight over the conduct of a Mr Big Operation through the rules of evidence.

Judicial gatekeeping

Australia evidence laws screen the admission of Mr Big confessions through threshold admissibility requirements, and judicial discretions to exclude improperly or unlawfully obtained confessions or to limit or exclude the use of unfair or unfairly prejudicial evidence (e.g., Evidence Act 2008 (Vic) ss 90; 135–138; or at common law through the fairness discretion R v Lee v (1950) 82 CLR 133, 49). Since the decision of the High Court majority in Tofilau, Australian courts have found that any prejudice flowing from the accused’s admissions and any apparent willingness to participate in criminal activities in scenario operations does not outweigh the probative value of the evidence (e.g., R v Weaven (No. 1) [2011] VSC 442). The majority approach has also been influential in courts’ rejection of arguments to exclude Mr Big confessions based on unfairness (R v Kilincer (No. 2) [2021] NSWSC 829, [42] (Johnson J); R v Rumsby [2023] NSWSC 229, [36] (Hulme JA)).

The common law threshold requirement of voluntariness is addressed through two ‘imperative’ rules: the basal voluntariness rule requiring proof on the balance of probabilities that the admission was made in the exercise of free choice, and the inducement rule (Tofilau v The Queen (2007) 231 CLR 396, 407, 411 and 469). Along with the right to silence, the protection of voluntariness recognises the ‘coercive power of the state’ (Tofilau v The Queen (2007) 231 CLR 396, 406 (Gleeson CJ)).

While the common law position is that confessions obtained by inducements held out by persons in authority are not admissible (McDermott v The King (1948) 76 CLR 501, 512 (Dixon J)), it is not entirely clear whether only inducements offered by persons in authority render confessions involuntary and inadmissible (Australian Law Reform Commission Report 26, 1985). For instance, Justice Dixon in McDermott v The Queen opined that ‘the basal principle that admissible confessions must be voluntary is not limited by any category of inducements that may prevail over a man’s will’ (McDermott v The King (1948) 76 CLR 501, 512). The High Court in R v Lee (1950) 82 CLR 133 also stressed that there seemed to be only one rule: ‘that statements must be voluntary in order to be admissible. Any one of a variety of elements, including a threat or promise by a person in authority, will suffice to deprive it of a voluntary character’ (144).

The rationale of the ‘person in authority’ requirement is to control coercive conduct by the state (Tofilau v The Queen (2007) 231 CLR 396, 495 (Callinan, Heydon and Crennan JJ)). Australian and Canadian courts have rejected arguments that ‘persons in authority’ included undercover police claiming to have access to corrupt officials whom the accused believed could influence the investigation and prosecution of the relevant crime (R v Grandinetti (2003) 178 CCC (3d) 449; Tofilau v The Queen (2007) 231 CLR 396, 406). In a minority judgement in Tofilau, however, Justice Kirby pointed out that state coercive power can also be employed on a suspect’s will by people who ‘appear to control the levers of state power’ but hold no public office (Tofilau v The Queen (2007) 231 CLR 396, [184]). Justice Kirby preferred the dissenting opinion in the Alberta Court of Appeal in Grandinetti:

… if undercover officers pretend to associate themselves with the police in such a way as to make it reasonable for an accused to believe that they can influence the prosecution, and an accused actually believes them, the officers can be persons in authority. (R v Grandinetti (2003) 178 CCC (3d) 449, 487 (Conrad JA))

Most Australian jurisdictions have adopted uniform evidence legislation and replaced the common law voluntariness tests with ss 84 and 85 Uniform Evidence Acts. Section 84 prohibits the reception of admissions obtained through oppressive conduct (R v Tang [2010] VSC 578, [25]) while the purpose of s 85 is to ensure that only reliable evidence is admitted (ALRC Report 102, [10.11]). Section 85 applies to admissions made in either of two circumstances: under s 85(1)(a) ‘to, or in the presence of, an investigating official who at that time was performing functions in connection with the investigation’ of an offence; or under s 85(1)(b) ‘as the result of an act of another person who was, and who the accused knew or reasonable believed to be, capable of influencing’ the prosecution of the accused (s 85(1)). Evidence of the admission is not admissible unless the circumstances in which the admission was made make it unlikely that the truth of the admission was adversely affected (s 85(2)).

Section 84 and 85 challenges to Mr Big admissions have been given short shrift by Australian courts (Deacon v The Queen [2019] NTCCA 21; R v Kilincer (No. 2) [2021] NSWSC 829; R v Rumsby [2023] NSWSC 229). In relation to s 84 for instance, Johnson J pointed out in Kilincer that the ‘very nature’ of the Mr Big technique involves pressure on the suspect to tell the truth ([2021] NSWSC 829, [148]). Other courts noted that covert operatives emphasise to Australian targets that their involvement with the organisation is voluntary (e.g., R v Cowan [2015] QCA 87, [85]–[87]; Deacon v The Queen [2019] NTCCA 21, [25]; Lauchlan v The State of Western Australia [2008] WASCA 227, [24]).

Only Kilincer and Rumsby have involved s 85 challenges, arguing that 85(1)(b) must apply to admissions made to covert operatives because the words ‘investigating official’ are absent in that subsection. The definition of the term ‘investigating official’ adopted in s 85(1)(a) specifically excludes covert operatives. The accused in Rumsby argued that it was inconsistent for the legislature to specifically exclude covert police from the operation of s85(1)(a) and not go on to specifically exclude covert operatives from the operation of s85(1)(b) (R v Rumsby [2023] NSWSC 229, [185]). Taylor has also asked ‘why should an admission made under circumstances which make it likely that its truth was adversely affected be admissible simply because no person in authority … was involved in its making?’ (2019, p. 55). The Explanatory Memorandum to the Evidence Amendment Bill 2008 (Cth), however, is clear that s 85(1)(b) amendments were intended to have the effect of excluding covert police despite the vagueness of the text ([114]). This interpretation was confirmed in R v Kilincer (No. 2) [2021] NSWSC 829 ([194] (Johnson J)) and R v Rumsby [2023] NSWSC 229 ([188] (Hulme JA)).

The Australian Law Reform Commission observed early that the narrow application of s 85 reflected ‘a broader concern’ about the potential for fabricated confessions and the impact this had on public confidence in the criminal justice system (ALRC, 2006, [10.13]). As the fallout of the Lawyer X scandal discussed below demonstrates, official deception and a cultural resistance to transparency and accountability have a devastating impact on public trust in the fairness, effectiveness, and integrity of policing processes. Amending s 85(1)(b) to include covert police as a person ‘capable of influencing the decision’ to prosecute under s 85(1)(b) would remove uncertainty by placing the onus on the prosecution to establish the requirements under s 85(2).

Additional scrutiny of Mr Big confessions may also be provided through appropriate jury directions.

Supporting evidence

Confessions are a category of potentially unreliable evidence (e.g., Evidence Act 1995 (NSW) s 165(1)(a)). Unless there are good reasons for not doing so, judges are required on request to warn the jury that evidence may be unreliable, inform the jury of matters that may cause it to be unreliable and warn the jury of the need for caution in deciding to accept the evidence (see e.g., Evidence Act 1995 (Cth), s 165(2)–(3)). Although s 164 of the uniform legislation abolished corroboration requirements, judges can invite juries to look for supporting evidence when giving an unreliability direction (e.g., Anile v The Queen [2018] VSCA 235, [202]). An example is the presence of non-public information in a suspect’s confession. Police have ‘strong practical reasons to test and to safeguard the reliability of a confession’ and to this end, may keep critical information from the public so that the admissions can be tested against matters only known to the police (Garrett, 2009, p. 1067). For example, in Kamalasan, the accused disclosed to Mr Big that he had poisoned the victim and how the poison had been administered. This disclosure was crucial when the police had kept the cause of death confidential (2019, [2], [34], [81]).

The Table 2 below details evidence relied on to support confessions in Australian Mr Big cases.

Table 2. Supporting evidence in Mr Big cases.

Type	Case	Evidence
Non-public information	Fesus	Fesus knew the location of his wife’s wallet and that she had been strangled when that information was not public knowledge ([174], [178])
	Karakas	Karakas provided new information that was not known to the police*
	Marks	Marks’ description of the use of the murder weapon could only have been known by the killer
	Kamalasanan	Kamalasanan admitted to poisoning the deceased with cyanide, which was only known to the medical examiner and the police (2019, [12], [14], [44])
	Tofilau	Tofilau told Mr Big he used a scarf to strangle the victim. A scarf was later found in that location
	Rumsby	Rumsby had ‘esoteric knowledge’ that the deceased’s pubic area was shaved
	Taylor	Taylor had actual knowledge that could only be known by the killer or who had communicated with the killer (2016, [98], [136])
Evidence of other witnesses	Favata	Substantial consistency between Favata’s admissions and statements of other witnesses ([46], [49])
	Ghiller	Admissions corroborated by numerous witnesses (2003, [56])
	Donai	Admissions confirmed co-accused’s evidence (2011, [71]).
	Standage	While Standage did not make ‘direct or express’ admissions, his statements tended to prove he ‘had means and motive’, corroborated the evidence of other witnesses, and were probative of tendency arguments and lies relied on by the Crown (2017, [105])
	Lauchlan	Listening device intercept of conversation after the confession confirmed admissions (2008, [50])
Subsequent confirmation by accused	Clarke	Clarke admitted in formal interview that he had smothered and probably stabbed the victim.
	Hill	Hill admitted in formal police interview and at trial that he had injured the victim, causing his death
	Bennett	Bennett was suspected in two other arson cases but there was little evidence of his culpability other than his admissions to covert police (2004, [25]). Bennett pleaded guilty
	Deacon	Deacon gave evidence at trial that he had killed the victim (2019, [27], [76])
Evidence derived from confession: leading police to remains or key evidence	Deacon	Deacon also led police to the victim’s remains (2019, [27], [76])
	Abrahams	Abrahams led police to the location of the victim’s remains (2013, [14])
	Cowan	Cowan led police to the proximate location of the victim’s body and clothing
	Jelic	Police found items from the victim’s wallet in the location indicated by Jelic to the covert operatives (2016, [4])
Inconsistent supporting evidence	Weaven	Weaven led police to the location of a knife. The knife differed from the description of the knife given to Mr Big and could not be established as the murder weapon (2011, [22]–[23]).
	Kilincer	Factual differences between Kilincer’s admissions and the forensic and ‘objective evidence’ including inconsistencies in description of murder weapon in ‘confession’ (2021, [175]).
	Simmons; Moore	Other evidence contradicted Simmons’ admissions about murder. Simmons was found not guilty of murder in a judge-only trial.

*It is unclear whether this information would have advanced the police case.

Non-public information included in confessions was the most significant type of supporting evidence occurring in seven cases (31.4%), followed by five cases (22.7%), in which the accused led police to real evidence, five cases (22.7%) in which the targets’ admissions were supported by other evidence, and four (18.18%) where the accused subsequently confirmed the admissions.

The prevalence of non-public or ‘holdback’ information is consistent with Iftene and Kinnear’s study of Canadian cases applying the Hart framework between 2014 and 2019 where they found that 44% of cases analysed involved confessions which included ‘holdback’ information (Iftene & Kinnear, 2020, p. 321). Knowledge of non-public information in a confession can provide strong support for its reliability (R v Marks (2004) 150 A Crim R 212). However, the cases do not really consider the possibility of contamination, namely that police may have inadvertently communicated non-public information to an accused prior to meeting Mr Big (Lutes, 2020, p. 218).

Similarly with Iftene and Kinnear’s study, all Australian Mr Big confessions have been admitted, even when there were inconsistencies in the supporting evidence such as in Kilincer, Weaven, and Simmons (338). Only Simmons was acquitted of an offence in a judge-only trial following the admission of a Mr Big confession, although he was found guilty of an earlier assault. This outcome may reflect the higher acquittal rate in judge-only trials than in jury trials (Small, 2021).

Key inconsistencies in the Kilincer and Weaven confessions related to the murder weapons used in each. Weaven was suspected of killing Mary Cook who was stabbed to death. During the Mr Big operation that followed, Mr Big told Weaven that the police suspected his involvement in Cook’s death, but the gang could ‘fix’ the situation in exchange for Weaven’s honesty. After Mr Big rejected two stories offered by Weaven, Weaven provided a third version in which he described killing Mary Cook with a screwdriver and a steak knife, which broke at the tip of the blade during the stabbing. He claimed the knife and broken piece were buried under a foot and half of soil and that he had thrown the screwdriver down a drain (Weaven, 2011, 2018). Weaven directed the police to where he had placed the knife and screwdriver, but the knife found by police did not match his description to Mr Big. The knife handle was visible above the ground and the blade was intact. Evidence from the autopsy was also equivocal, namely that it was ‘reasonably possible that the knife had not caused the wound’. The screwdriver was never located (R v Weaven (No. 1) [2011] VSC 4, [23]). Justice Weinberg considered that these discrepancies could be explained without rejecting the confession (R v Weaven (No. 1) [2011] VSC 442, [42]).

Kilincer was suspected of killing his employer and became the target of a Mr Big operation 24 years later following a cold case review. The deceased had been found lying next to a bloodied sledgehammer in his workshop with severe head injuries and a metal pipe in his mouth. Kilincer eventually made a confession to Mr Big that he killed his employer with metal pipes, but several features of the confession were inconsistent with the physical evidence. Kilincer had denied using a sledgehammer when that question was put to him by Mr Big, and he claimed he used two metal pipes in the attack while only one metal pipe was found at the scene. Justice Johnson found, however, that the inconsistencies could have been attributable to the passage of time (R v Kilincer (No. 2) [2021] NSWSC 829, [235]).

In Simmons, the suspect’s confession that he had assaulted, and later murdered, Andrew Russell was admitted despite numerous inconsistencies between that account and other evidence. In a trial by judge alone, Justice Hamill, however, accepted it was possible that Simmons lied to ‘big-note’ himself and concluded that the undercover operation ‘had the capacity to induce a false and unreliable confession’ ([482]). He found Simmons not guilty of murder but guilty of assault. Simmons is the only example of an acquittal in an Australian Mr Big case The acquittal rate in judge-only trials in NSW has increased over the last 15 years as judge-only trials have become more widespread (Croucher & Hon, 2022, p. 84), which might, in part, explain the different outcome in Simmons.

Glazebrook suggested that Mr Big confessions should not be treated differently from other types of evidence that are known to be fallible (2015). We argue, however, that unless Mr Big operations are subject to rigorous independent scrutiny, Mr Big convictions should be supported by the availability of a specific warning that addresses susceptibility to making false confessions. While we recognise there is a need to simplify and streamline jury directions, this does not mean that new, clear directions should not be added to address issues of particular importance.

‘Trickery and deception’

The problem of balancing the need for police to protect their operatives and methodologies with the need for transparency and open justice has been addressed in several applications for suppression orders (e.g., R v Jelicic (Review of Suppression Orders) [2016] SASC 58; R v Simmons (No. 5) [2015] NSWSC 333). While details of the technique are published in media and legal websites, and the technique has ‘its own Wikipedia entry’ ([2015] NSWSC 333, [42] (Hamill J)), evidence of Mr Big operations is still frequently curated to protect investigative methodologies and the identities of covert operatives.

The need for deception and secrecy raises questions about the adequacy of oversight regimes to protect against over-zealousness (Loftus et al., 2016, p. 644). This deficit was highlighted by the 2020 Royal Commission into the Management of Police Informants in Victoria, which inquired into the registration and use of a Victorian barrister Nicola Gobbo (Lawyer X) as a police informant. Victoria Police failed to disclose that Gobbo gave information to the police about people who were or had been her clients. Gobbo’s conduct breached her ethical duties to her clients, violated legal professional privilege, and tainted numerous convictions.

The Royal Commission found evidence of ‘systemic failure’ in Victoria Police (RCMPI, 2020, p. 22), observing signs of a police culture where officers were driven to get results at any costs; a culture of exceptionalism leading to internal attitudes that police can operate outside the rules; poor ethical decision making; limited consideration of human rights; and a resistance to exposure, accountability and transparency. The extent to which Victoria Police privileged getting results at the expense of rights and the proper administration of the criminal justice system indicates that it is time to reconsider the current assessment of Mr Big confessions.

Evidence of Mr Big operations is limited by disclosure decisions made by the police about the relevance of covert recordings because not every covert recording made will be perceived as directly relevant to the guilt or innocence of the accused by the police (Ruyters et al., 2021). Further, while communications during covert operations are recorded under surveillance device warrants (e.g., Surveillance Devices Act 1999 (Vic) s 15), these may not necessarily record conversations occurring prior to the commencement of the recording or after the recording is complete when key facts may have been suggested to the target (Garrett, 2009, p. 1079).

While full disclosure is fundamental to a fair trial (Grey v The Queen [2001] HCA 65), police in Australia and the United Kingdom have exhibited a longstanding cultural resistance to this golden rule (Ruyters et al., 2021). The very late disclosure by police in 2023 of up to 530 h of secret recordings at an inquiry into Kathleen Folbigg’s convictions for murder reveals that this is an ongoing problem (Whitbourn, 2023). As Justice Hamill indicated in Simmons, not every transcript of recordings disclosed may necessarily be read (R v Simmons; R v Moore (No. 2) [2015] NSWSC 143, [43]–[45]). Justice Hamill himself discovered a crucial conversation that had not been transcribed and errors in the identification of speakers in some covertly recorded conversations. While Justice Kirby noted in Tofilau that the confessional evidence secured through the scenarios was ‘apparently reliable, in the sense that it was accurately recorded’ (Tofilau v The Queen (2007) 231 CLR 396, [146]), this assumes adequate disclosure.

Conclusion

It is easy to be complacent about Mr Big operations in Australia. Police use of this tactic is likely to attract public sympathy because of associations with the convictions of notorious child murderers Brett Cowan and Matthew Clarke. More significantly, the technique has been approved by the High Court in Tofilau.

Mr Big confessions are a last resort for police. Confessions justify the significant costs of running such operations, decisions to launch investigations, and the emotional and psychological toll on law enforcement participants (Kowalczyk & Sharps, 2017). On the other hand, occupational pressures to get results, the lack of external oversight, and police secrecy exceptionalism (Koningisor, 2023) may create a dangerous potential for serious miscarriages of justice. Mr Big operations raise clear risks that suspects may be induced into making false confessions. Suspects may be incentivised to falsely confess to gain assistance with perceived police suspicion, to benefit financially by becoming a full ‘gang’ member, or merely to ‘boast’ and gain esteem amongst high-ranking gang members.

Comprehensive reviews of legal frameworks relating to the use and oversight of covert investigative powers in Australia are urgently needed, particularly in the context of Mr Big operations which are an especially intrusive police investigative method. There is a particular need for enhanced transparency and rigorous accountability mechanisms. Current oversight regimes are inconsistent and confusing, while existing external oversight for covert activities such as phone tapping and surveillance is limited to compliance with the legislative regimes under which they are established and authorised.

We argue that Mr Big operations that do not involve controlled activities should be brought under controlled operations frameworks. Assessments about the reliability of Mr Big confessions should be assisted by including covert police within the scope of evidentiary rules about inducement-based confessions, explicitly within the scope of s85(1)(b) Uniform Evidence Act. Finally, we argue that juries should receive appropriate direction addressing susceptibility to making false confessions.

Disclosure statement

The authors work with the Bridge of Hope Innocence Initiative, which investigates wrongful conviction claims.