That funny feeling in Division 105A of the Criminal Code: Benbrika and the INSLM Review
19.11.2024
On 19 December 2023, Abdul Nacer Benbrika was released from custody after being detained for the best part of 20 years. He was released after Hollingworth J of the Victorian Supreme Court made orders replacing a continuing detention order (CDO) with an extended supervision order (ESO) pursuant to div 105A within Pt 5.3 of the Criminal Code (Cth) (Code). Division 105A is a Commonwealth legislative scheme of ‘post-sentence orders’ (PSOs) whose object is:
… to protect the community from serious Part 5.3 offences by providing that terrorist offenders who pose an unacceptable risk of committing such offences are subject to [a CDO or ESO].
As reported, Hollingworth J’s judgment was ‘scathing’ of the Commonwealth’s failure to disclose a research report which seriously undermined the validity of a risk assessment tool used by the Commonwealth’s experts to establish the ongoing risk posed by Mr Benbrika. Her Honour has referred the matter to the Independent National Security Legislation Monitor (INSLM) for further investigation, describing the non-disclosure as a ‘serious interference with the administration of justice’ under what is already ‘very unusual and draconian legislation’ and observing (at [316]):
What happened in this case should never have happened, and should not be repeated in the case of Mr Benbrika or any other person the subject of a post-sentence order application.
This piece will examine the judgment of Hollingworth J in Benbrika v Attorney-General (Cth) [2024] VSC 265 (published 5 June 2024) alongside the earlier review of the 4th INSLM into div 105A (published on 3 March 2023). Both documents contain vital critiques of the scheme, and my contention is that they raise real questions as to whether the stated object of community protection, so fundamental to the constitutional validity of div 105A, is really being pursued by the Commonwealth; or whether, to go further, it is an unattainable and therefore misleading target.
A. Background
Part 5.3 of the Code
This Part of the Code deals with terrorism offences. A ‘serious Part 5.3 offence’ carries seven or more years of imprisonment. A PSO can be made when a person has been convicted of a serious Part 5.3 offence (s 105A.3(1)(ii)), is at least 18 years old (s 105A.3(1)(c)) and is in custody (s 105A.3A). The application for a PSO (either a CDO or ESO) is brought by the ‘AFP Minister’ in the Supreme Court of a State or Territory (s 105A.5). A court considering an application for a PSO can: (a) make a CDO under s 105A.7; (b) make a ESO under s 105A.7A; or (c) dismiss the application. A CDO results in the relevant ‘offender’ being detained in a prison for the duration of the order (s 105A.3(2)), whereas an ESO means that the relevant offender will live in the community under conditions (s 105A.7B).
Section 105A.6B sets out matters to which the court must have regard, including ‘any report of an assessment received from a relevant expert’ (sub-s (1)(b)). To make a CDO, the court must be satisfied ‘to a high degree of probability … that the offender poses an unacceptable risk of committing a serious Part 5.3 offence’ and there is ‘no less restrictive measure’ which could be imposed (s 105A.7(1)). To make an ESO, the court must be satisfied ‘on the balance of probabilities’ of the same ‘unacceptable risk’ and that the conditions to be imposed are ‘reasonably necessary, and reasonably appropriate and adapted, for the purpose of protecting the community from that unacceptable risk’. Importantly, community protection on the basis of risk assessment is the raison d’être of div 105A.
Mr Benbrika
In September 2008, Mr Benbrika was convicted by jury of three offences under Pt 5.3 relating to his leadership of a terrorist organisation. He was sentenced to 15 years’ imprisonment with a non-parole period of 12 years (R v Benbrika & Ors [2009] VSC 21). Mr Benbrika was not granted parole and his sentence was due to expire on 5 November 2020. Before this could occur, the Minister for Home Affairs (Minister) made an application to the Victorian Supreme Court for a CDO. Tinney J granted this application on 24 December 2020 and made a CDO for three years to 23 December 2023 (Minister for Home Affairs v Benbrika [2020] VSC 888). On 13 December 2022, Mr Benbrika applied for review of the CDO; on 6 March 2023, the Commonwealth applied for an ESO in place of the CDO. These two applications were considered together by Hollingworth J (heard in June 2023).
Over this period, Mr Benbrika initiated two appeals which ended up in the High Court. The first was a challenge to the constitutional validity of CDOs: Minister for Home Affairs v Benbrika [2021] HCA 4 (Benbrika I). The argument advanced was that an order for continuing detention under div 105A impermissibly conferred non-judicial power on a court exercising federal judicial power under Chapter III of the Constitution, contrary to the separation of powers and to the principle set out in Chu Kheng Lim v Minister for Immigration [1992] HCA 64 at [23]:
… putting to one side the exceptional cases … the involuntary detention of a citizen in custody by the State is penal or punitive in character and, under our system of government, exists only as an incident of the exclusively judicial function of adjudging and punishing criminal guilt.
A majority of the Court upheld the validity of div 105A. Kiefel CJ, Bell, Keane and Steward JJ held at [40] that ‘Division 105A has an evident non-punitive, protective purpose’ tied to ‘the assessment of the risk the offender poses of future harm to the community upon release and not on punishing the offender for the offence for which he or she was sentenced’. Their Honours also noted at [12] that the power to make a CDO ‘is subject to a number of statutory safeguards’, including the Commonwealth’s duty to make inquiries and disclose information which would tend against the making of an order (see s 105A.5(2A) and (3)(aa)).
Mr Benbrika’s second visit to the High Court was because, in November 2020, the Minister cancelled his Australian citizenship. And then, on 12 May 2023, a few weeks before the trial before Hollingworth J, the Minister cancelled Mr Benbrika’s ex-citizen visa. However, on 1 November 2023, the High Court held in Benbrika v Minister for Home Affairs [2023] HCA 33 (Benbrika II) that s 36D of the Australian Citizenship Act 2007 (Cth), which created a ministerial power to cancel the citizenship of dual citizens convicted of particular offences, was invalid as it purportedly reposed in the Minister the exclusively judicial function of punishing criminal guilt. Mr Benbrika’s citizenship was thus restored.
The INSLM and the Corner Report
The INSLM is an independent, statutory reviewer. Grant Donaldson SC, the then INSLM, carried out a review into div 105A and public hearings were held on 22-23 June and 21 November 2022. The review was published on 3 March 2023. (Unless otherwise indicated, references to the INSLM are to Mr Donaldson SC as officeholder.)
A focus of the review was on risk assessment and the role of the ‘relevant expert’. Mr Donaldson SC explained that to provide a framework to whichever expert is retained by the Commonwealth (usually a clinical psychologist), the Department of Home Affairs (Department) is a licensee (the sole licensee in Australia) of a tool called Violent Extremism Risk Assessment – Version 2 Revised (VERA-2R) (at [268]). The tool is a quasi-scientific, quasi-actuarial framework said to facilitate ‘structured professional judgment’ – a questionnaire of sorts which goes into the expert report produced to the court to lend credibility to what is a clinical judgment by the expert as to the risk of an individual committing a terrorist act (or, specifically, a respondent’s risk of committing a Pt 5.3 offence).
In the November hearing, a research report surfaced: ‘Testing the Reliability, Validity, and Equity of Terrorism Risk Assessment Instruments’ by Dr Emily Corner and Dr Helen Taylor (Corner Report). The Corner Report aimed to gauge the efficacy of VERA-2R, and another tool called Radar. The Department commissioned the Corner Report in 2018 (Hollingworth J at [254]) and it was provided to the Department in May 2020. Mr Donaldson SC first became aware of the existence of the report when he visited Dr Corner’s webpage (at [270]). This prompted him to issue a compulsory notice to produce to the Department. The Corner Report was provided to him as part of the review around September 2022. The Corner Report’s findings included that VERA-2R and Radar were based on theoretical and empirical evidence of ‘poor quality’; and that VERA-2R factored in only a small proportion of statistically relevant variables, offered ‘poor predictive validity’ and was unable to predict the ‘specified risk with anything other than chance’ (see the Executive Summary of the Corner Report).
The Corner Report remained suppressed by the Department until obtained by the INSLM and acknowledged on the public record in November 2022; this explains Mr Benbrika’s application for review of his CDO the following month.
Overview of the judgment of Hollingworth J
Her Honour had two applications to determine: Mr Benbrika’s application for review of the existing CDO and the Commonwealth’s application that the CDO should be replaced with an ESO. Mr Benbrika ‘accepted the reality’ that it was open to the Court to impose an ESO on the basis of an ongoing risk (at [59]), without formally conceding the issue. The primary dispute between the parties was the length and conditions of any ESO.
In assessing whether Mr Benbrika posed an unacceptable risk, Hollingworth J turned to the ‘specified matters’ a court must have regard to (at [75]), including the expert evidence. The Commonwealth’s expert was Dr Chelsey Dewson who assessed Mr Benbrika’s overall risk as ‘moderate-low’ (at [72]). Mr Benbrika’s expert was Dr Anne Speckhard who described the risk as ‘miniscule’ (at [104]). Her Honour concluded that while Mr Benbrika had made steps towards deradicalisation, as at December 2023 he ‘posed an unacceptable risk of radicalising others to commit acts of religiously motivated extremism’ (at [131]).
Her Honour reasoned then through the conditions that should be included in the ESO (at [133]-[244]). The balance of the decision, at [245]-[322], dealt with the ‘non-disclosure issues’ by the Commonwealth of a number of reports (including the Corner Report) which undermined the legitimacy of tools including VERA-2R. This is elaborated upon below.
Mr Benbrika was ordered to be subject to an ESO for the duration of one year, rather than the three years sought by the Commonwealth. This was partly because, in her Honour’s view, the Commonwealth should not continue ‘to have the benefit of the improper non-disclosure’ (at [329]) as there was ‘no doubt that the administration of justice has been adversely affected’ by the non-disclosure (at [326]).
B. That funny feeling
Reading the judgment of Hollingworth J alongside the observations of the INSLM, one cannot escape the funny feeling that, despite judicial oversight, div 105A is being administered in a way contrary to the stated object of community protection so fundamental to its constitutional validity as found by a majority in Benbrika I. There is also a legitimate question as to whether the level of risk of an offender committing a serious Pt 5.3 offence is ascertainable to any accurate or meaningful standard.
I will turn first to the judgment of Hollingworth J on the non-disclosure issue and other conduct of the Commonwealth, and then the wider reaching conclusions of the INSLM. It is worth foregrounding that the key recommendation of the INSLM review is that CDOs should be abolished.
Hollingworth J on the non-disclosure of the Corner Report and other shortcomings of the Commonwealth
The AFP Minister has an obligation under the Code to disclose information which suggests that a CDO or ESO should not be made (s 105A.5(3)(aa)). Her Honour was of the view that the Commonwealth had clearly failed to comply with this safeguard via non-disclosure of the Corner Report (and other similar reports) (at [24]-[25] and [298]). To be clear, the Corner Report was sitting on a shelf in the Department when the original CDO was made in relation to Mr Benbrika at the end of 2020, and the validity of the risk assessment process was a key issue in those original proceedings. A decision had apparently been made by the Department to suppress the Corner Report (at [263]).
Hollingworth J considered evidence filed by the Commonwealth accounting for the non-disclosure. Her Honour accepted the explanation and apology provided by AGS lawyers (at [247]). In contrast, her Honour rejected that provided by the Department as ‘utterly disingenuous’ in circumstances where the Department ‘has deliberately chosen not to put a proper explanation before the court’ (at [299]). The judgment at [250]-[304] examines the unsatisfactory affidavit evidence filed on behalf of the Department in a manner reminiscent of the Robodebt Royal Commission. Internal correspondence and file notes paint a frightening picture of ignorance or incompetence. ‘No affidavit was filed on behalf of anybody in Home Affairs who had personal knowledge of any of the relevant meetings, communications or decisions’ which Hollingworth J inferred was a deliberate attempt at obfuscation (at [250], [297]-[298]). Her Honour also found that persons within the Department appreciated how damaging the Corner Report could be; and one correspondence showed the researchers being asked to ‘rewrit[e] the report in a way more favourable to Home Affairs’ (at [259]).
Other shortcomings of the Commonwealth can be noted. At a crucial pre-trial juncture (May 2023), the Minister cancelled Mr Benbrika’s ex-citizen visa; ‘no explanation has been given … as to why the cancellation decision was made at that particular time’ (at [39]). Had Mr Benbrika’s citizenship not been restored in Benbrika II, the cancellation would have rendered the question in front of Hollingworth J practically moot because ‘release’ under an ESO would mean only that Mr Benbrika moved into immigration detention (rather than detention under a CDO). Further, on 15 May 2023, a delegate of the Minister decided that any immigration detention would take place in a correctional setting, namely Long Bay gaol in NSW, refusing to approve any less secure placement which kept Mr Benbrika in Victoria (nearer his family and lawyers) (at [41]-[42]). Hollingworth J was clearly troubled by this, noting that the Commonwealth’s own application before her was for Mr Benbrika to return to life in the community on conditions. Her Honour was also critical of witnesses called by the Commonwealth: on the question of risk, Mr Benbrika’s expert had ‘vastly more experience’ (at [120]), and on the question of ESO conditions, the Commonwealth’s two witnesses presented a ‘mantra-like repetition of the statutory test’ (at [139]-[147]).
Hollingworth J observed that the obligation to disclose ‘exculpatory material is a fundamental safeguard to ensure the protection of individual liberty’ (at [315]). Together, these failings seriously undermine the confidence vested in the Commonwealth, in its varied emanations, to fairly and justly administer the PSO scheme. At best they are reflective of incompetence; at worst, they are symptoms of a detention-at-all-costs, punitive mindset which is entirely at odds with the espoused statutory object of community protection.
The INSLM’s questioning of the basis of the div 105A scheme
Mr Donaldson SC stated that he was ‘shocked’ by the non-disclosure of the Corner Report (at [290]) and was concerned about ‘a very great risk of abuse of the process of applying for such orders’ (at [336]), noting that the scheme cannot be justified by trusting ‘those in power’ to act ‘properly’ or only when ‘necessary’ (at [301]). However, if the latest Benbrika case illustrates the maladministration of div 105A, the INSLM review went a step further by challenging the basis of the scheme: community protection from an unacceptable risk. The tenor of the INSLM review is captured at [3]-[4]:
… these laws have made us a coarser and harsher society. I doubt that anyone knows whether they have made us safer.
In gentler times these kinds of laws did not exist, although some deny it. In times past, people who had committed crimes and served their sentences were thought to have paid their debt to society and lived with us in our communities, even though some reoffended.
Mr Donaldson SC stated ‘[r]isk assessment is the heart of the Div 105A regime’ and the ‘role of the relevant expert … is central’ (at [240]). Reports such as the Corner Report represent an existential threat to schemes like div 105A because they suggest that the exercise of risk assessment is one not capable of being carried out. Chapter 7 of the review calls into question the legitimacy of ‘risk assessment for violent extremist offending’ as a legitimate body of expertise. Mr Donaldson SC asked (at [259]):
What if there is not, and could never be, coherent, validated evidence of the risk of a person committing extremist violence? Should Div 105A then continue as it is, with people being detained or having their freedom curtailed based on predictions that profess but lack validity?
Or using the definition of ‘relevant expert’ in s 105A.2, what if there is no one ‘competent to assess the risk of a terrorist offender committing a serious Part 5.3 offence’? Importantly, ‘a protean field of research’, based on a pool too minute to support statistical analysis (at [253], [303]), ‘should not form a substantial part of the basis for deciding whether a person is detained in custody in a prison’ (at [294]).
Finally, Mr Donaldson SC considered it ‘profoundly telling, when considering whether the CDO mechanism is proportionate to the risk of terrorism in Australia’ that neither the United Kingdom nor New Zealand had any comparable scheme despite the terrorist events which have befallen those countries and not Australia (at [309]). Yet, in Australia, there has been a lack of any real resistance to these laws by the Parliament or public, ‘although unease may be drowned out by the enthusiasm and rancour of proponents’ (at [325]). Mr Donaldson SC expressed his own feeling of disquiet about the scheme as follows (at [340]):
The regard in which the Australian judiciary is held, and the sprinkling of judicial pixie dust on this power to order detention in a prison, should not obscure an irrebuttable risk of injustice. The risk of error posed by untestable judicial predictions about future behaviour, with the consequence of error being that a person will be detained in a prison for no good reason, cannot be ignored.
C. Conclusion
The Parliamentary Joint Committee on Intelligence and Security is currently reviewing the operation, effectiveness and implications of div 105A, following the INSLM review. The Government has yet to respond to the review and its far-reaching recommendations, including the abolition of CDOs.
As for Hollingworth J’s referral of the Department’s conduct in the latest Benbrika proceeding to the INSLM for further review, the current INSLM, Jake Blight, in a statement of 19 June 2024 indicated that as his functions concerned law reform rather than the investigation of individual conduct no further action could be taken by the office. However, following consultation between the Department, the Attorney-General’s Department and the Public Service Commissioner, it was decided that there will be an independent preliminary inquiry ‘to determine whether there should be a code of conduct or other process in relation to the non-disclosure of material by any officer in the Benbrika proceedings’.
Of course, the non-disclosure of material like the Corner Report affects not only Benbrika but every proceeding under div 105A since at least May 2020. And, in the meantime, we are left with the funny feeling that div 105A is far from the gold standard for community protection from terrorism accepted by the High Court in Benbrika I. Generally, there is a tension between Mr Donaldson SC’s observations about the risk of preventive detention schemes being abused by those in power and the need for the Court to assume, at least on questions of constitutionality, that the objects and provisions of a law will be faithfully followed by administering agencies.
Foucault observed in Discipline and Punish that modern criminal justice functions and justifies itself only by grabbing onto other things; by dressing punishment up as something else. The issue here, however, is far from being merely philosophical; it is about the troubling reality of the State’s power to detain people on what appears to be conjecture, in the name of community protection.
Samuel Naylor is studying a Masters in Comparative Law at the Sorbonne. Before that, he worked as a lawyer in Broome.
Suggested citation: Samuel Naylor, ‘That funny feeling in Division 105A of the Criminal Code: Benbrika and the INSLM Review’ (19 November 2024) <https://www.auspublaw.org/blog/2024/11/that-funny-feeling-in-division-105a-of-the-criminal-code-benbrika-and-the-inslm-review>