Attorney General of New South Wales v AB (a pseudonym)

By summons dated 22 August 2025, the plaintiff (“the Attorney General”) seeks an order against the defendant pursuant to ss 121, 127(1)(a) and 128 of the Mental Health and Cognitive Impairment Forensic Provisions Act 2020 (NSW) (“the Forensic Provisions Act”) for the extension of his status as a forensic patient for a period of two years from the date of the order.

The defendant opposes the order, contending that the Court ought not be satisfied to a high degree of probability of the matters required by s 122(1) of the Forensic Provisions Act and, in the alternative, even if the Court is so satisfied, that the Court should exercise its discretion against making an order.

On 19 November 2025, following a preliminary hearing on 28 October 2025, Dhanji J made orders for the examination of the defendant under s 126(5) of the Forensic Provisions Act and an interim order under ss 130 and 131 for the extension of the defendant’s status as a forensic patient for a period of three months from 22 November 2025: State of New South Wales v AB (a pseudonym) (by his tutor Ramjan) (Preliminary) [2025] NSWSC 1357.

Ms Zipparo examined the defendant by AVL and furnished a report to this Court on 12 January 2026. Dr Elliott examined the defendant by AVL and furnished a report to this Court on 13 January 2026.

In addition to those reports, the plaintiff relies on the following material:

The defendant relies on the following material:

The defendant, a 45 year-old man, has a lifelong history of intellectual disability and cognitive impairment, apparently related to hypoxic brain injury at birth. He has one sibling, an older sister, but he has not had contact with his father for 15 to 20 years and ceased contact with his mother and sister from 2019.

The defendant left school in year 11. According to the agreed facts, there is “file information” (the content of which is not before the Court) that the defendant left school due to behavioural difficulties including sexually inappropriate behaviour. The defendant worked in gardening and landscaping intermittently, and has received the disability support pension since he was 16. He has had support and funding from the National Disability Insurance Scheme (“NDIS”) since 2020.

The defendant has had an intimate relationship with a woman described variously in the evidence as the defendant’s “friend”, “ex-partner” and “girlfriend” since about 2018. I propose to describe her in this judgment as the defendant’s partner. While the evidence suggests that this relationship may have been a closer one before the defendant’s most recent period of incarceration, it is clear that there is a current relationship which involves the defendant speaking to the partner daily and seeing her fortnightly (they live some distance apart). On the evidence, the partner is a significant source of support to the defendant.

The defendant has been convicted of driving offences on four occasions, including most recently in 2007, being drive while disqualified and drive with middle range prescribed concentration of alcohol. He was sentenced to 12 months’ imprisonment. The defendant’s other criminal history prior to the index offending includes the following convictions:

The defendant’s criminal history also includes the following charges which were dismissed pursuant to s 32 of the Mental Health (Criminal Procedure) Act 1990 (NSW) (or Mental Health (Forensic Provisions) Act 1990 (NSW), as that Act was titled from 2008 until its repeal, upon the enactment of the Forensic Provisions Act):

In relation to the first two sets of dismissed charges, I have treated the evidence of the defendant’s admissions with real care given the defendant’s intellectual disability and the fact that I do not have the whole of the prosecution brief for each offence. Furthermore, while the 1997 ERISPs record that the defendant (himself a vulnerable person) had a support person with him, the evidence in this Court does not disclose whether he had a support person with him when he was interviewed in 2003. Nevertheless, although I have treated these admissions with real care, I have taken that evidence into account when considering these allegations of sexual misconduct, a matter to which I return below.

On 23 October 2020, the defendant was charged with five counts of intentionally causing a fire being reckless as to its spread. The offences, which were found proven on the limited evidence available, involved the following:

On 20 October 2021, Herbert DCJ found that the defendant was unfit to be tried for the index offences and that he would not become fit within 12 months.

On 27 Oct 2022, Coleman DCJ found that, on the evidence available, the defendant had committed the offences. On 13 July 2023, Coleman DCJ imposed a total effective limiting term of 2 years and 18 days, commencing on 12 August 2021 and expiring on 30 August 2023.

On 24 August 2023, following a preliminary hearing under the Forensic Provisions Act, Wright J made an interim extension order extending the defendant’s status as a forensic patient for three months: Attorney General for New South Wales v Andrews (by his tutor Johnson) (Preliminary) [2023] NSWSC 1059.

On 22 November 2023, Weinstein J made an order extending the defendant’s status as a forensic patient for a period of two years: Attorney General for New South Wales v Andrews (by his tutor Johnson) (Final) [2023] NSWSC 1468.

When the Mental Health Review Tribunal (“the Tribunal”) reviewed the defendant in February 2024 and again in July 2024, it determined that the defendant should remain detained in a correctional centre.

On 27 February 2025, the Tribunal granted the defendant conditional release. The Tribunal was of the view that the defendant would require comprehensive support to live safely in the community. The defendant was released from custody on 19 March 2025 and has been in the community since that date.

The Tribunal reviewed the defendant on 11 June 2025 and 15 August 2025. Following the review on 15 August 2025, following evidence that the defendant’s living conditions were not acceptable or safe for him, the Tribunal varied the defendant’s accommodation condition to provide for the defendant to reside at his now-current address in Gorokan, NSW.

The defendant’s current NDIS plan, which is for three years to 20 November 2028, includes:

By reason of his status as a forensic patient, the defendant has received support and supervision from the Community Safety Program (“CSP”), including his case manager Mr Lindshield and team leader Mr Wu. The CSP is a program of Corrective Services NSW which provides case management of forensic patients with cognitive impairment in the community. Through the CSP, the defendant has also been referred to the Treatment and Rehabilitation Centre (“TRC”), which is part of the NSW Justice Health and Forensic Mental Health Network. The defendant engaged with the TRC between 27 March 2025 and 30 May 2025. On 15 July 2025, the TRC discharged the defendant for reasons related to “ambivalence … about engaging in treatment for problem sexual behaviour”. However, the defendant was referred to the TRC a second time on 5 December 2025 and was accepted on 10 December 2025. While the defendant experienced some distress during the first sessions (apparently due to questions about the defendant’s sexual functioning and relationship with his partner, which the defendant regarded as intrusive), the evidence before me indicates that the defendant and his support staff all agree on the utility of the TRC treatment.

The defendant started attending Alcoholics Anonymous (“AA”) meetings upon release from custody. While the defendant either has ceased or appears likely shortly to cease attending those meetings (as well as having disengaged for a period in September/October 2025), that is in the context of having demonstrated a sustained commitment to abstinence. This commitment to abstinence is a matter in the defendant’s favour. I address it further below.

Part 5 of the Forensic Provisions Act provides for forensic patients (among others) and Part 6 provides for the extension of a person’s status as a forensic patient. Section 69 provides that the objects of Parts 5 and 6 are:

Section 122 of the Forensic Provisions Act sets out the test for when an extension order can be made:

The standard of proof, “high degree of probability”, is higher than the civil standard but lower than the criminal standard: Attorney General (NSW) v SK (Final) [2025] NSWSC 1173 (“SK”) at [14] per N Adams J, citing Cornwall v Attorney-General (NSW) [2007] NSWCA 374 at [21] in relation to the same language in similar legislation, the Crimes (High Risk Offenders) Act 2006 (NSW).

The phrase, “unacceptable risk” should be given its everyday meaning in its context, having regard to the objects of the legislation: Lynn v State of NSW (2016) 91 NSWLR 636; [2016] NSWCA 57 (“Lynn”) at [58]. It is directed to the assessment of risk in the context of making the community secure from harm as opposed to guaranteeing its safety and protection: Lynn at [61]; see also SK at [15]. The assessment is by reference to both the likelihood of the risk eventuating and the gravity of its potential consequences: Attorney General (NSW) v Randall (Preliminary) [2023] NSWSC 226 at [24].

“Serious harm” may concern physical or psychological harm. In Attorney General (NSW) v Kereopa (No 2) [2017] NSWSC 928 at [16], RA Hulme J said:

As to the second limb, whether the risk can be adequately managed by other less restrictive means, this requires the Attorney General to prove the negative to the relevant standard: Minister for Mental Health v Paciocco [2017] NSWSC 4 at [8]; SK at [19].

Whether alternative means are more or less restrictive is to be judged by the legal power of others to control the defendant’s actions, locations, treatment and other matters, as well as the practical operation of how that power might be exercised in a particular instance: Attorney General (NSW) v Doolan (No 2) [2016] NSWSC 107 at [96]; Attorney General (NSW) v McGuire [2019] NSWSC 76 at [29].

The Court is not confined to asking whether there are alternative means available which “could” in theory manage the risk, without regard to whether those means would in fact be effective to manage the risk. In Attorney General (NSW) v Haines (No 2) [2025] NSWCA 248 at [88], the Court of Appeal (Adamson, McHugh and Ball JJA) held that such an approach:

An unacceptable risk is “adequately managed” if it is mitigated by the proposed management regime so that the community’s interest in being kept safe is outweighed by the community’s interest in not having mentally ill or mentally disordered individuals or forensic patients being confined in some form of institutional care rather than taking their place in the community: Attorney General v McGuire (No 2) [2014] NSWSC 288 at [63].

An application for an extension order may only be made in respect of a forensic patient if they are subject to a limiting term or an existing extension order: s 124(1) of the Forensic Provisions Act.

An application may not be made more than 6 months before the end of the limiting term or the expiry of the existing extension order, as the case may be: s 124(2).

An application must be supported by documentation that addresses each of the matters referred to in s 127(2), to the extent relevant to the application: s 125(a). Pursuant to s 125(b), the documentation must include a report prepared by a qualified psychiatrist, registered psychologist or registered medical practitioner:

In determining whether to make an extension order, s 127(2) provides that the Court must have regard to the following matters in addition to any other matter it considers relevant:

The Court has a discretion as to whether or not to make an extension order, and the interests of a defendant may be taken into account in the exercise of that discretion: Attorney General (NSW) v Peckham (Final) [2019] NSWSC 1775 at [15] and [122].

I am satisfied that the threshold requirements in ss 124 and 125 of the Forensic Provisions Act are met.

The defendant was, at the time of the commencement of these proceedings on 22 August 2025, subject to an existing extension order. These proceedings were commenced less than 6 months before the expiry of that extension order. Section 124 is satisfied.

The application was (and is) supported by documentation which addresses each of the matters in s 127(2) to the extent relevant, thereby satisfying s 125(a). The documentation includes a risk assessment report by Dr Youssef dated 16 March 2025 which satisfies s 125(b).

I have placed significant weight on the safety of the community.

I have read carefully the reports of the two Court-appointed experts. Both experts also gave evidence at the final hearing. I return to the experts’ evidence in their reports and at the hearing, below.

It is appropriate to address an issue that arose at the final hearing which had the potential to impact the weight that might be placed on the experts’ reports (although, as I explain, in the context of the present case that potential was not realised).

The two Court-appointed experts were briefed with an agreed five-volume bundle of documents that included, behind tab 1, Justice Dhanji’s reasons and orders following the preliminary hearing. Dr Elliott in his report stated that he had read Justice Dhanji’s reasons closely, and at a number of places identified “conclusions” from those reasons to which he had regard in preparing his report. As emerged from their oral evidence, both experts said that Justice Dhanji’s reasoning or conclusions played a part in the formation of their opinions.

I do not doubt that his Honour’s reasons were included in the experts’ briefs with good intentions – among other things, they record the reasons for those very experts being appointed – but doing so gave rise to a real risk.

When a judge makes orders following a preliminary hearing under s 126, the judge is making an interlocutory and in some respects preliminary assessment based on whether the matters alleged in the plaintiff’s documentation would, if proved, justify making an order. The Court is not generally weighing up the documentation, predicting the ultimate result, or considering what evidence a defendant might call: Attorney General (NSW) v Tillman [2007] NSWCA 119 at [98]. While the Court does assess the “weight” of particular aspects of evidence as well as the “weight” of the supporting documentation as a whole (State of NSW v Heness (Preliminary) [2019] NSWSC 1710 at [29]), the Court does so for the particular purpose of a preliminary hearing. The test has been described as being akin to that which formerly applied to the determination of whether a prima facie case existed sufficient to justify the committal of a defendant to stand trial, and may involve taking evidence at its highest: State of NSW v Rosenburg (Preliminary) [2025] NSWSC 191 at [21]. The Court may give particular weight to risk-avoidance: Attorney General (NSW) v Winters [2007] NSWSC 611 at [7]. A defendant will typically be slow to put on any evidence in advance of a preliminary hearing, and such evidence may have limited utility on the primary question for the Court (which is directed to the matters alleged in the plaintiff’s documentation). A situation should not be created in which defendants perceive an indirect strategic benefit in putting on evidence and litigating issues at a preliminary hearing because whatever appears in the judge’s reasons might directly or indirectly influence any Court-appointed experts.

The facts which a judge decides to emphasise in a judgment following a preliminary hearing are for the specific purpose of giving reasons for the orders being made. Such reasons may be very brief and are not generally prepared with a view to the experts relying upon them. It is not the function of such reasons to give a comprehensive conspectus of the evidence likely to be relevant at final hearing – indeed, such reasons might not even identify what the judge considers to be the most salient features of a defendant’s risk. Rather, the reasons may focus on what was the subject of dispute or debate at the preliminary hearing. There is a risk of unfairness to a defendant if an expert’s attention were drawn more specifically to certain issues merely because they were the subject of dispute or debate at the preliminary hearing.

I appreciate that reasons following a preliminary hearing sometimes contain a convenient summary of the evidence before the Court at that hearing, and experts may consider that they are assisted by that summary. Dr Elliott said as much in his evidence to me. However, even where reasons summarise the salient evidence before the Court at the preliminary hearing, there is an unavoidable risk that the judge may summarise the evidence, or may be perceived to summarise the evidence, in a manner that is informed by the judge’s own (preliminary) views of the substantive merits. Experts’ opinions should not be influenced directly or indirectly by the qualified risk assessment exercise a judge conducts following a preliminary hearing. Steps should also be taken to avoid the risk of a perception of influence.

Dr Elliott and Ms Zipparo in their oral evidence before me both said that they had regard to Justice Dhanji’s reasons and that in formulating their opinions they were influenced by (or benefitted from) his Honour’s reasoning and conclusions. I make no criticism of either expert for doing so. It no doubt reflected their respect for this Court and the opinions of its judges, and the Court acknowledges that respect.

I received oral evidence from each expert separately on this topic before receiving evidence concurrently as to their substantive opinions. I was satisfied, following questions asked by the Attorney General’s counsel and additional questions asked by me, that each expert considered themselves able to express an independent opinion without regard to Justice Dhanji’s reasons, although Ms Zipparo candidly acknowledged that this is sometimes difficult. Each expert confirmed in their separate oral evidence that their opinions would be the same in all respects without regard to Justice Dhanji’s reasons. Neither party made any submission to the contrary.

I am satisfied that both experts were in fact able to confirm their opinions without placing any weight on Justice Dhanji’s reasons. Indeed, I should not be taken to suggest that the inclusion of preliminary hearing reasons in an expert’s brief is likely to impact adversely an expert’s report or opinion. Experts subscribe to the Expert Code of Conduct and “[i]t would be a rare case that allegations would risk causing an experienced independent medical practitioner to be distracted from conducting an objective assessment of the defendant’s medical condition”: Attorney-General (SA) v Rowe [2025] SASC 189 (“Rowe”) at [109] per B Doyle J. In this case, I accept each expert’s evidence that their opinions would not be any different had they not been briefed with those reasons. I have not qualified the weight I place on the experts’ opinions by reason of this matter.

As I have said above, I make no criticism of the legal representatives or the experts, but with the benefit of hindsight and the experience in this case, it would be preferable not to include preliminary hearing reasons in briefs to Court-appointed experts under this statutory regime or closely analogous statutory regimes. If potentially relevant matters emerge from the reasons that the parties agree should be highlighted for the experts, or if there is a convenient and neutral summary of the facts that the parties agree would be of assistance to the experts, then that can be included in the letter of instructions, or as an annexure to the letter of instructions. Similarly, if there are particular matters that the judge at a preliminary hearing considers should be drawn to the experts’ attention (such as, where current criminal charges are pending, the need to take care about a defendant’s right to silence), those matters can also be raised in the letter of instructions rather than by providing the experts with the judge’s reasons. Of course, the Court’s reasons will generally be published and therefore be in the public domain. The Court need not prevent experts from reading such reasons if they choose to do so – as B Doyle J observed in Rowe the experts can generally be trusted to form their own independent opinions – but it is preferable not to include the reasons in the experts’ briefs, lest that imply that the experts are being invited to read and place weight on those reasons.

I would add in that regard that the material before me discloses a high level of cooperation between the parties’ legal representatives, including on matters such as the preparation of the briefs to the experts and the constructive approach of both sets of legal representatives to the concerns I raised once I had raised them. While this is necessary to efficient performance of the Court’s function and reflects the fact that these are civil proceedings subject to the guiding principles in Part 6 of the Civil Procedure Act 2005 (NSW), it is also to be commended. Proceedings of this kind are complex, involve a high volume of material, and must be decided in limited time periods. The Court recognises and appreciates the cooperative and constructive approach of parties’ legal representatives in such litigation.

I have read carefully Dr Youssef’s risk assessment report dated 16 March 2025, as well as her two addenda dated 13 August 2023 and 24 August 2025 respectively. I received oral evidence from Dr Youssef at the hearing, which I have also considered carefully. I address that evidence where relevant below.

The defendant submits that Dr Youssef’s risk assessment should be treated with care because her opinions are based in part on charges that did not result in convictions, and because her March 2025 risk assessment report erroneously said that the defendant’s “last recorded sexual offence” occurred in 2017. The defendant submits that the only proven sexual offence by the defendant was committed in 2007. During her oral evidence, Dr Youssef said that she had been referring to the “peep and pry” offence, and that she should have referred to a charge and not a recorded offence. While I recognise that the “peep and pry” offence was withdrawn, I can understand why one might describe an offence as “sexual offending” when it involved unlawfully being outside the window of a dressing room, late at night, while a woman was dressing in the room. Regardless of whether one labels the conviction as “sexual offending”, I am satisfied that it is offending of a kind that is material to the defendant’s risk of sexual offending in the future.

The evidence before me also included two earlier reports prepared by Dr Youssef in relation to the defendant, a report dated 6 August 2023 (in the context of the 2023 extension proceedings) and an addendum report dated 13 August 2023. I have considered those reports, although they have primarily been useful only as context, given that Dr Youssef’s current opinions are reflected in her more recent reports and her oral evidence.

The evidence before me included a number of earlier reports by psychiatrists, psychologists or medical practitioners, including the risk assessment report and the reports of the two Court-appointed experts before Weinstein J in 2023. I have read and considered all of the reports in the evidence before me, but I have been mindful of Dr Youssef’s helpful recognition in her oral evidence before me that there is a real potential for risk assessments from 2023 to have become inaccurate or unreliable with the passage of time.

Furthermore, the experts before me specifically disagreed with some of the views expressed by Dr Furst in his report, including in particular Dr Furst’s opinion that the defendant needed to remain as a forensic patient indefinitely (ie, lifelong). As Ms Zipparo pointed out, in addition to the passage of time, there have been material changes including the defendant having an opportunity to develop a stable lifestyle with all the supports currently available to him. Given Dr Furst’s opinion was based on the evidence before him, and given the passage of time and the changed circumstances, I have not placed weight on Dr Furst’s opinion that, as at November 2023, it appeared that the defendant might need to remain as a forensic patient indefinitely.

The Tribunal granted the defendant conditional release and, in doing so, was satisfied that this would not pose a risk to the safety of the defendant or others. However, this was in the context of a case management plan and involved the imposition of conditions to enable that plan to be implemented.

I have considered the Tribunal’s orders and reasons following review hearings since the imposition of a limiting term. The Tribunal was satisfied following the hearing on 27 February 2025 that conditional release was the least restrictive care that was available, but this was in the context of recognising a moderate risk of reoffending, the importance of abstinence, and the need for the defendant to receive comprehensive support. The release was on the basis of a plan put forward by the CSP, and the plan from the CSP was plainly an important feature of the Tribunal’s assessment.

When the matter came before the Tribunal on 11 June 2025, 16 August 2025 and 10 February 2026, the Tribunal continued to order release but subject to conditions. I note in particular, in the Tribunal’s reasons following the hearing on 11 June 2025, that the “feedback from the treating team and [the defendant’s] support workers was universally positive” and that the defendant had made the transition from custody to the community well.

The evidence before me includes a number of letters or reports from the CSP. I have considered that material.

Most significantly, I have a letter prepared by Mr Lindshield, senior clinical consultant with the CSP and the defendant’s case manager, on 19 December 2025 and co-signed by Mr Wu, team leader, on 21 January 2026. That letter records, among other things, the following:

As to the defendant’s attendance at AA meetings, I note the defendant’s own evidence that he did not cease attending for two months following a confrontation, but rather he missed four meetings (the meetings being weekly) in about September or October 2025 following the confrontation with the other participant. Little turns on this difference in the circumstances, but I assume in the defendant’s favour that he only missed four meetings.

Similarly, there is some lack of clarity on the evidence about whether the defendant is in fact performing any landscaping at present – he told Dr Elliott that he had all the necessary equipment but (at least as at the date of his interview with Dr Elliott on 17 January 2026) he had yet to secure any clients. However, this does not adversely affect my assessment. The weight of the evidence indicates that the applicant has occasional employment through his own business and is looking to develop it further.

These matters are very much to the defendant’s credit. They are among the ways in which the defendant has made significant progress in the community. However, as I explain further below, I am unable to conclude that the defendant’s rehabilitation, alcohol abstinence treatment or employment prospects have yet reached a stage where they will be a reliable source of stability into the future. I address below the evidence about what additional steps warrant consideration to reduce the defendant’s risk.

I have received into evidence and reviewed more detailed reports and case notes from the CSP and the TRC, all of which predate this correspondence. That material is consistent with the letter from Mr Lindshield of 19 December 2025, countersigned by Mr Wu on 21 January 2026.

On the evidence, the defendant has complied with his obligations as a forensic patient. This is consistent with the positive progress the defendant has made in the community. I have taken it into account in the defendant’s favour.

I am mindful that, if the 2014 allegations are true, the defendant reoffended sexually while still on parole for his 2009 sexual offending. I take this into account, but I place little weight on it, given it was over a decade ago and the allegations are unproven.

I have considered carefully the reasons of Coleman DCJ (as his Honour then was) when he imposed the limiting term on 13 July 2023 and of Weinstein J when he imposed the extension order on 30 November 2023.

Justice Weinstein set to one side Dr Furst’s opinion as to the defendant’s risk of future fire-setting or drink driving, because his Honour was satisfied that the defendant posed an unacceptable risk of causing serious harm to others on the basis of the risk of sexual offending alone. Dr Youssef, Dr Seidler and Dr Furst all recognised the risk of sexual reoffending. Indeed, his Honour concluded, “on the totality of the evidence [before his Honour], in my view the court could come to no other conclusion.”

I am conscious that these conclusions were based on the evidence then available, and that the defendant has made material progress in the community since that time. Nevertheless, I have taken these views into account in making my assessment.

In addition to reports from psychologists, psychiatrists, medical practitioners and government agencies, which I have taken into account under the relevant subparagraph of s 127(2) above, the evidence includes reports from other professionals, which I take into account under s 127(2)(i).

From April 2025 to August 2025, the defendant received NDIS funding support for a mental health occupational therapist. I have read a report from that therapist to the Tribunal. The author of that report expressed some concerns about the defendant’s current intimate relationship, on the basis that it may hold some unhealthy patterns, but the author recorded that the defendant is open to engaging in discussions about this and is accepting of challenges to his thinking. The report also records that the defendant accepts the need to maintain his own interests and quality of life outside of the relationship.

That report was in the context of the defendant having an unsatisfactory NDIS provider in relation to accommodation and support services. As I explain below, the evidence indicates that after August 2025, when the defendant changed his accommodation and his NDIS provider, the defendant has developed a very good working relationship with his current NDIS provider and support workers. Unfortunately, the defendant has not had the benefit of a mental health occupational therapist since August 2025 because his previous NDIS funding ran out. However, on the evidence before me, the defendant has further funding for this to resume, following the commencement of the defendant’s current NDIS plan in November 2025.

The defendant also places weight on a longitudinal report in respect of the period October 2025 to February 2026 prepared by Ms Michelle Bray, an advanced behaviour support clinician, who provides support funded by the NDIS.

Ms Bray’s report includes the following:

I received expert evidence during the hearing that the preparation of a report like this – which is one of the things for which the defendant receives funding under the NDIS – has a protective function, because it is an overall progress report. I return to the protective value of the NDIS support further below.

I have placed some weight on Ms Bray’s report, but I found it significantly less pertinent to the Court’s task under s 121 of the Forensic Provisions Act than the expert reports of Dr Youssef, Dr Elliott and Ms Zipparo. It was prepared for a different purpose, and appeared to be directed towards whether the Tribunal should order involuntary medical treatment. Furthermore, the evidence does not disclose Ms Bray’s qualifications nor the material with which she was briefed. Ms Bray has not undertaken to be bound by the Expert Code of Conduct. None of this is a criticism of Ms Bray – it is clear that she was preparing a report for a different purpose, not with a view to giving evidence as an expert in these proceedings.

I have set out above the defendant’s criminal history, including charges that were dismissed under s 32 of the Mental Health Act. It is common ground that the Court may have regard to alleged criminal conduct, even if it did not result in a criminal conviction: State of NSW v O’Donnell (Final) [2018] NSWSC 563 at [9]. As the plaintiff points out in submissions, dismissal under s 32 of the Mental Health Act does not involve a finding that the defendant did not commit the offences, or even that the prosecution was unable to prove them.

I have taken care to bear in mind that such allegations were not proved against the defendant. In relation to the 2014 allegations, I received into evidence the transcripts of interviews with the two alleged victims. I have taken the same approach to those transcripts that Adamson J took in Attorney General (NSW) v Steadman [2016] NSWSC 174 at [27] and [124]. That is, I have taken them into account with respect to the pattern of offending conduct, but have significantly qualified the weight I have placed on that evidence, since they were not tested at trial or in any other context.

Furthermore, while I have taken into account in the defendant’s favour his completion of Self-Regulation Programme: Sexual Offending, an intensive custodial sexual offending program for men with an intellectual disability or cognitive impairment, I have borne in mind that the defendant is alleged to have reoffending sexually in 2014, and reoffended in 2017 in a manner which, while not necessarily “sexual offending”, is nevertheless a matter I regard as relevant to the defendant’s risk of sexual reoffending. Even if the defendant did reoffend sexually after completing the custodial program, that would be consistent with Dr Elliott’s evidence that he was not convinced that such programs would significantly reduce the defendant’s risk, and Dr Youssef’s evidence that certain kinds of targeted programs will be of less value for a person with the defendant’s intellectual disability and cognitive impairment.

The evidence before me clearly establishes that, were the defendant to reside unsupervised in the community without any structured supports, he would pose an unacceptable risk of causing serious harm to others.

Ms Zipparo diagnosed an Intellectual Development Disorder (Intellectual Disability – Mild), characterised by impairments to cognitive functioning, including impaired verbal comprehension, extremely poor memory interfering with his ability to learn new skills, and impairment of executive function meaning that he has poor planning, mental flexibility and decision-making abilities, and a tendency to be impulsive. To similar effect, Dr Elliott diagnosed a mild intellectual disability characterised by impairments in executive function, impairments in short-term memory, and difficulties learning new skills and with social skills.

Both Court-appointed experts also diagnosed Alcohol Use Disorder, which both experts recognised to be in “genuine remission”. Dr Elliott noted that substance use disorders are frequently characterised by periods of abstinence followed by relapses at times of stress or other triggers, although the pattern is less frequent with more sustained periods of abstinence and greater resilience to stress. Dr Elliott stressed that the defendant experienced stress in his previous accommodation but managed to remain sober, and reported that the defendant “stated emphatically that he will not return to alcohol use should he encounter difficulties”.

Dr Elliott and Ms Zipparo both considered that the defendant poses a risk of causing serious harm to others if he ceases to be a forensic patient and is not subject to any other type of order. The focus of the risk for both experts was sexual reoffending, although the experts also recognised a risk – albeit one that is difficult to quantify or offer a structured professional judgment about – of further fire-setting offending. Dr Elliott described the risk in oral evidence as “low”, but neither Ms Zipparo nor Dr Youssef gave evidence that the risk was low. In any event, the authorities are clear that a low risk can be unacceptable, depending on the circumstances. For myself, taking into account all the evidence before me, I would not characterise the risk as low. Care is often needed in approaching descriptions of a risk as “low”, “medium” or “high”, as this can mean different things in different contexts and I do not propose to use those expressions myself. For the reasons set out herein, I consider (and am satisfied to the requisite standard) that the risk is unacceptable.

On the evidence, factors that would materially increase or contribute to this risk would include significant life stressors, a lack of suitable report, or a return to alcohol use. The experts agreed that the risk posed by the defendant is low if the defendant continues to receive the high levels of support from which he is currently benefitting, and if he continues to enjoy stability in his current intimate relationship.

These risks are exacerbated by the defendant having only a limited social support network. His relationship with his partner is clearly protective (I return to it below) and there is some evidence of other socialising, such as with his neighbour, but the defendant does not have a significant social support network. While I have taken into account his close relationship with his NDIS support workers, described by him as “friends” and even “family”, I am also conscious of Ms Zipparo’s evidence that the defendant has poor boundary-setting with his support workers. While the defendant’s engagement with his NDIS support network is a significant matter in his favour – as I develop further below – I am nevertheless satisfied that the absence of a significant social support network is relevant to the defendant’s risk and has a tendency to increase it.

I have taken into account evidence before me that the risk of sexual offending declines the longer individuals with a history of sexual offending remain sexual-offence-free in the community. In particular, I received evidence from research (Hanson et al, “Reductions in Risk Based on Time Offense-Free in the Community” (2018) 24(1) Psychology, Public Policy, and Law 48) and from both Court-appointed experts that after 10 to 15 years, most individuals with a history of sexual offending are no more likely to commit a new sexual offence than individuals with a criminal history that does not include sexual offences, although with “very high risk” offenders, that period “stretches out to around 20 years”. While I have taken this evidence into account, I cannot conclude that the defendant has been sexual-offence-free in the community for 10 to 15 years, mindful that (a) he has been in custody for some of the last 10 to 15 years, (b) while the 2017 offending might not warrant the label “sexual offending” the conduct may well have involved that kind of offending, and (c) while the 2014 allegations of sexual offending were not proved, I have evidence including an ERISP from both alleged victims which I have taken into account in the manner explained above.

The defendant has made considerable progress in the community since his release on 19 March 2025. This is very much to his credit.

He has remained abstinent, bringing his period of abstinence to well over three years (albeit including some time in custody). He is rightly proud of that achievement. He has shown significant insight into the risks associated with alcohol, including that it can get him into trouble and could lead to him ending up in gaol. There is no indication of the defendant associating, or seeking to associate, with people who are drinking or are likely to facilitate his drinking. These are all protective factors that significantly reduce the risk of relapse.

The defendant has also demonstrated resilience in the face of material stressors on at least two occasions. The first was the defendant’s response to the stress caused by his previous accommodation arrangements. On the evidence, they were plainly unsatisfactory (indeed, insanitary). Despite finding the situation distressing, the defendant remained sober and continued to engage with his supports. Later, the defendant experienced stress in his AA meeting, as noted above, when tension arose between him and another participant. The other participant suggested that the defendant was an undercover policeman which, as Dr Elliott explained, is a particularly distressing accusation to have levelled for a person who has recently been incarcerated. Again, although the defendant responded with a level of stress and anxiety, the defendant did not relapse or cease to engage with his supports. These are important matters which demonstrate that the defendant is developing resilience and prosocial coping strategies.

The defendant’s intimate relationship with his partner is also a source of support and a protective factor. I have set out at [8] above what I am able to infer from the evidence about this relationship. As noted at [72] above, the defendant’s mental health occupational therapist (when he received that service) expressed some reservations about aspects of this relationship, but not in a manner that would undermine my conclusion that it has a material protective effect. However, I need to take into account the risk that this relationship may cease, or may go through a difficult period. I do not accept the defendant’s submission that the possibility of the relationship ending is mere pessimistic speculation. The fact that all relationships are at risk of coming to an end does not mean the risk is any less real. While the relationship is of some years’ standing, the evidence indicates that its nature has fluctuated. Mr Lindshield, in an email to the defendant’s GP in December 2025, expressed some concerns about the longevity of the relationship or the extent of the commitment. While I have no detail about the reasons for Mr Lindshield’s concerns, I am left in a position where I cannot be confident that the relationship will continue indefinitely. The evidence is clear that if it ceases, that could lead to a major increase in the defendant’s risk profile.

I have referred above to the evidence of the defendant’s employment in his own business cleaning/detailing cars and landscaping. This is another example of the progress the defendant has made, although the experts – especially Ms Zipparo – considered that the defendant’s risk would be better addressed if he were more fully employed. The defendant gave evidence that he had been to APM Employment Services on 14 January 2026 and had met with an employment agency officer who said they could help him find work with companies that hire people with a disability. The defendant says he is keen to get work. If this occurs, it will provide a further protective factor.

The defendant currently receives significant support funded by the NDIS, summarised at [21] above. He has clearly benefitted substantially from this support and has developed an excellent relationship with his current provider. While I can never exclude the risk of the NDIS funding and support changing materially, nor the risk of the defendant deciding not to accept or engage with that support, the evidence before me indicates that the support is likely to continue at least for the duration of the defendant’s plan, and that the defendant is likely to continue to accept and engage with that support.

His NDIS support includes a detailed behaviour support plan prepared by Ms Bray, which I have reviewed and accept has a protective effect. On the expert evidence before me, that plan is at least in part directed towards assisting the support workers. I have also taken into account Ms Bray’s longitudinal summary of behavioural data, addressed above.

I accept the evidence that the NDIS support is a material protective factor – indeed Ms Zipparo described it as “a very protective factor”. It includes the defendant receiving daily support, which provides a kind of supervision, albeit not of the same nature or extent as the supervision he receives as a forensic patient. The defendant’s NDIS support workers are likely to see, or be able to tell, if the defendant resumes drinking or starts engaging in other risky behaviours. However, as the Court-appointed experts acknowledged in their evidence before me, NDIS support workers have no coercive powers – at most they can refer them to other services, such as drug and alcohol treatment or a mental health service. There is a significant prospect that this would not prevent the risky behaviours once they started. While the defendant’s positive engagement with his NDIS support providers is a valuable matter in his favour, it would not on its own be adequate to manage the defendant’s risk as matters currently stand.

The defendant’s engagement with the CSP, and his recent reengagement with the TRC, is also very much to the defendant’s credit.

While it is to the defendant’s credit that he wishes to engage with the TRC, the evidence on this topic if anything reinforces the importance of the defendant’s forensic patient status. As a result of that status, the defendant is supervised and supported by the CSP team. They have been able to make an assessment – conscious of the risks posed by the defendant – about the relative benefits of AA and TRC. That assessment may change over time. If the defendant experiences a destabilising event or ceases to be willing to engage constructively with TRC, then the defendant’s unwillingness to engage with AA may loom larger in the defendant’s risk profile. If the defendant ceases to be a forensic patient, this support and supervision would cease.

If the defendant ceases to be a forensic patient, the defendant will continue to have access to the NDIS support, as well as receiving informal support from his partner. He will have access to AA, but will cease to have access to the CSP and, it appears, the TRC. It that situation, there would be grounds for optimism, but not confidence, that the defendant would maintain his impressive period of abstinence and his pursuit of further employment. I am unable to conclude that, as matters presently stand, the NDIS support is a substitute for the support and degree of oversight provided by the CSP, or that the NDIS support adequately addresses the defendant’s risk without the additional support of the CSP and the additional supervision that comes with being a forensic patient.

The expert witnesses were invited to identify what else they would consider necessary or desirable to address the defendant’s risk. The two concrete interventions proposed concerned regular psychological treatment and employment.

As to regular psychological treatment, Dr Youssef and Ms Zipparo both considered that this would have a significant impact on the defendant’s risk profile, if he engaged positively with it. Dr Elliott, when asked whether he saw a need currently for the defendant to engage with a psychologist on a regular basis, said, “It would be helpful, yes”. Related to this, Dr Youssef identified the importance of the defendant internalising the skills necessary to self-regulate, something that may take some time given his intellectual disability and cognitive impairment.

As to employment, the evidence from Mr Lindshield and from the defendant indicates that the defendant is taking active steps towards further employment, although some of this has occurred only quite recently. The evidence is clear that if the defendant has stable, full-time employment, then that is likely to have a significant favourable impact on his risk profile. The evidence is that he is likely to need significant support in any full-time employment, but the evidence also indicates that such support is presently available to him and appears likely to continue. Nevertheless, the defendant does not presently have stable full-time employment, and the evidence does not enable me to say whether or when that will occur.

Additional matters, emphasised particularly by Dr Youssef but also discernible in the evidence from Ms Zipparo, are the importance of the defendant internalising the skills necessary to self-regulate, and the benefit of developing a wider social network or wider engagement with the community. I recognise that these matters may follow from the first two concrete proposals (regular psychological treatment and employment) if those proposals are fully implemented, although that would be a matter that may need to be assessed at a later time.

Despite the existence of protective factors and the defendant’s creditable progress in the community, I am satisfied to a high degree of probability that the defendant poses an unacceptable risk of causing serious harm to others if his status as a forensic patient is not continued. While I consider that the greatest risk relates to sexual reoffending, I do not exclude the risk of other kinds of offending, including fire-setting and drink-driving, each of which are capable of causing serious harm to others.

I am satisfied to the requisite standard that the defendant would pose this unacceptable risk even taking into account the other supports and protective factors which would remain if the defendant ceased to be a forensic patient. An alternative way to consider all the other supports and protective factors is that the proposed management regime, absent forensic patient status, would not mitigate the defendant’s risk such that the community’s interest in being kept safe is outweighed by the community’s interest in not having mentally ill or mentally disordered individuals or forensic patients being confined in some form of institutional care rather than taking their place in the community: Attorney General v McGuire (No 2) [2014] NSWSC 288 at [63]. In reaching this view, I recognise that the legislation does not require the elimination of all risk before a person might cease to be a forensic patient. However, I am satisfied that the risk that would be posed by the defendant, if the defendant ceased to be a forensic patient but continued to receive the other support and structure presently available, would be an unacceptable one.

As submitted by the defendant and accepted by the Attorney General, the Court retains a discretion whether or not to order the extension of a person’s forensic patient status, even if the Court is satisfied to a high degree of probability of the matters in s 122(1): see also [37] above. The defendant submits that the Court should exercise its discretion against making an order, having regard to the defendant’s progress in the community and the supports that are in place and likely to continue, particularly the NDIS supports. I have taken those matters into consideration.

I have not been persuaded to place weight, as the Attorney General invited me to, on a perception that the defendant would himself benefit from extension of his forensic patient status. The defendant submitted that this is paternalistic. There is scope for real debate about whether the Court should make its own judgment about whether a defendant’s personal interests are better served by continuation of forensic patient status, at least where the defendant appears through an independent tutor and experienced counsel. It is not necessary for me to resolve this issue, as I am comfortably persuaded to exercise my discretion in favour of making an order without taking this perception into account. Accordingly, while I appreciate the point of view that forensic patient status enables certain services and supports to be provided which are in the defendant’s evidence, I have not placed weight in the exercise of my discretion on that contention.

I am satisfied that it is appropriate for the Court to exercise its discretion in favour of making an extension order.

The Attorney General sought an order for 18 months. The defendant contended that, if an order is made, it should be for the shortest possible period (nominating six months as the period if the defendant was not successful in his primary position that no order should be made).

Dr Elliott’s view was that 12 months was a sufficient period for any extension order. Ms Zipparo favoured an order for 18 months, but that was on the basis that it may take up to 12 months to put the additional systems in place and then six months to test the robustness of those systems. The “additional systems” that Ms Zipparo had in mind were further employment and regular engagement with a psychologist. Ms Zipparo said that if it only took six months to put those systems in place (and I do not see why it would take longer than six months), then the order would only need to be for 12 months. Dr Youssef recommended an extension for 18 months to two years.

I have taken into account that, if the order is for only 12 months, then the Attorney General is likely to start taking steps to consider an application for a further extension order in a few months’ time. I recognise this creates the prospect of the Attorney General needing to take material steps in circumstances where the defendant may be progressing well, and where the preliminary steps for another application under s 123 of the Forensic Provisions Act could be its own source of stress to the defendant. Indeed, I am conscious that if the extension order is for only 12 months, then in light of Ms Zipparo’s and Dr Youssef’s evidence there is a distinct possibility – perhaps likelihood – that a further extension application will be made, although of course the merits of such an application will fall to be assessed by reference to the evidence and circumstances at that future time. It is to be hoped that those supporting the defendant will assist the defendant to understand that if another extension application is made, that may be partly referable to the decision today to order an extension for 12 months rather than 18 months.

As a general matter, the considerations advanced by the Attorney General will often provide a proper basis for making an order for a longer time. However, given the particular evidence before me, I consider that it is in the interests of both the safety of the community and the defendant to have a slightly shorter period of time during which the defendant can work towards demonstrating his reduction in risk. Given the defendant’s progress in the community to date and the evidence that if this continues the defendant may cease to pose an unacceptable risk in about 12 months, I am satisfied that an order of no more than 12 months’ duration is warranted.

I make the following orders: