< Return to Contents | Previous | Next >
Operation Yamaha: description and assessment
- In this chapter we outline what happened on Operation Yamaha. We address the allegations that Qari Miraj was assaulted by New Zealand Special Air Service (NZSAS) personnel shortly after he was captured, and was tortured while in the custody of the Afghan National Directorate of Security (NDS) to make a confession. We then give a brief description of relevant developments in relation to New Zealand’s detention policy after Operation Yamaha, before undertaking our assessment in light of the Terms of Reference.
Operation Yamaha: the facts
- The following account of what happened on Operation Yamaha is based on the oral and documentary evidence available to us. In terms of the lead-up to the operation, it incorporates findings from the Inspector-General of Intelligence and Security’s (IGIS) Inquiry.
- There was high-level New Zealand Government interest in the activities and whereabouts of the insurgents involved in the 3 August 2010 attack on the New Zealand Provincial Reconstruction Team (NZPRT) in Bamyan province in which Lieutenant Tim O’Donnell was killed. Government Communications Security Bureau (GCSB) personnel in Wellington worked with New Zealand Defence Force (NZDF) personnel in Afghanistan and the deployed New Zealand Security and Intelligence Service (NZSIS) staff to identify specific members of the group involved in the attack as their priority “High Value Targets”. One of these was Qari Miraj, a well-known insurgent operating out of Baghlan province.
- Miraj had been trained in the construction of improvised explosive devices (IEDs) and had a number of insurgent fighters under his command. He was considered to have been responsible for the deaths of Afghan security personnel in Tala wa Barfak and was a leading member of the group that carried out the 3 August 2010 attack. He was the person who had likely exploded the IED that destroyed the vehicle in which Lieutenant O’Donnell was travelling. Miraj was sought by Afghan authorities, and Task Force 81 (TF81) had been looking to capture him for some time. Miraj was a dynamic target, in the sense that he was often on the move.
- Operation Yamaha was one of four deliberate detention operations over the period of the Operation Wātea deployment to detain the high-risk insurgents identified as having planned or conducted attacks against the NZPRT. Two of the other deliberate detention operations were Operation Burnham and Operation Nova. The purpose of Operation Yamaha was to disrupt the insurgent network based in the Tala wa Barfak district of Baghlan province, which was planning attacks against the NZPRT, Afghan forces and local people in Bamyan province. The evidence does not indicate that the operation was motivated by a desire for revenge.
- In August 2010, shortly before Operation Burnham, Miraj was placed on the Joint Prioritised Effects List (JPEL), with TF81 listed as the lead agency. In September 2010, the Commander of the NZPRT contacted TF81 to express his concern about the threat posed by Miraj and to ask TF81 to consider mounting an operation against him. TF81 did undertake some preliminary planning in September 2010 for an operation to capture Miraj in a village in Tala wa Barfak where he had a compound, but ultimately that operation did not go ahead. Obviously, that was an operation which, like Operation Burnham, would have required the Chief of Defence Force’s consent as it was outside Kabul and its surrounding districts. We also note that TF81 learnt in mid-September 2010 that Miraj had just passed through Kabul, probably going either to or from Pakistan, but too late to attempt to capture him.
- Moving forward to Operation Yamaha in January 2011, we have described the planning process for the International Security Assistance Force (ISAF) operations in which TF81 was involved with Afghan partners in the context of Operation Burnham and will not repeat that discussion here. The planning process for Operation Yamaha was somewhat attenuated. On Saturday 15 January 2011, the NZPRT received information that helped to identify and locate Miraj in Kabul. This information was passed on to TF81, which carried out a range of checks before mounting a “short notice” operation.
- Shortly before the time of his expected arrival, TF81 obtained approval from ISAF and from the Afghan Government for a night-time deliberate detention operation to capture him. Arrangements were made to deal with deconfliction issues, and to arrange for the availability of an airborne surveillance asset. TF81’s usual Afghan partner was the Afghan Crisis Response Unit (CRU). We heard evidence that the CRU was not available for this operation, but Afghan National Directorate of Security (NDS) personnel were involved, along with some Afghan National Police (ANP). In particular, the NDS provided an investigator and a prosecutor equipped with a warrant for Miraj’s arrest. The IGIS’s report indicates that the deployed NZSIS staff managed TF81’s relationship with the NDS, both generally and in relation to this operation. It records that an NZSIS officer’s view on the NDS as a partner was that “it was a reliable and dependable partner with a professional prosecution system”.
- Initially the plan was to conduct a roadside detention at a location where there were no civilians around. However, that changed at the last moment when it became clear that Miraj would be staying the night in Kabul, and the plan became to enter a building and capture Miraj there. There was no time to issue formal orders. They were written in a note book. But the rules of engagement applied, which, of course, limited the use of force.
- Soon after midnight on 16 January 2011, some TF81 support personnel went from TF81’s base camp to the compound where Miraj was thought to be staying. A covert cordon was deployed to the area. About two hours later, TF81 ground forces and support personnel left TF81’s base camp to go to the staging area at Kabul International Airport, where they met the personnel who had located the compound. They were at the staging area for around one and a half hours, during which time the Ground Force Commander completed his planning (including risk assessment and operational plan) and conveyed his orders to the troops and back to base camp. The Afghan partner forces arrived and, shortly after 4am, TF81 and Afghan partner forces moved to the general area of the compound.
- On arrival in the general area, the New Zealand and Afghan forces moved in on foot and some TF81 personnel set up a cordon near the compound. NZDF documents from the time record that once the cordon was set, ANP and NDS personnel conducted a “soft knock” at the compound. When there was no response, TF81 personnel entered the compound by going in over the outside wall (a “covert manual entry”). They searched the compound, apparently in conjunction with the NDS personnel, although they did not conduct a tactical site exploitation (that is, the gathering of critical information and material from the site).
- While the compound was occupied, it turned out that Miraj was not there. It was then realised that he was in fact in a building on the other side of the road. That building was a mosque, which non-Afghan forces could not enter due to cultural sensitivities. Accordingly, TF81 personnel established a cordon at the mosque and ANP and NDS personnel conducted a “soft knock” entry. Once they had entered the mosque, they found Miraj and four companions, all of whom surrendered peacefully. One of those with Miraj was another high value target. While in the mosque, ANP and NDS personnel conducted a tactical site exploitation.
- There were no arrest warrants for Miraj’s four companions, and we understand that two of them were ultimately released at the scene. It is not clear to us whether Afghan personnel formally arrested Miraj inside the mosque or whether that occurred outside on the road. In any event, Miraj was quickly identified by TF81 personnel, as he had a distinctive appearance and they had photographs of him. TF81 personnel began to process the five men. This involved “plasti-cuffing” them, searching them for weapons and such like, and removing items such as mobile phones so that data could be stripped from them. While several witnesses thought some biometric processing (such as taking fingerprints and retina scans) was carried out on the road outside the mosque, the weight of the evidence was that it was carried out later at the NDS facility before TF81 personnel handed the detainees over to NDS personnel there. Given that evidence and the fact that the nature of the road-side locale made taking biometric data difficult, we think it probable that the biometric processing was carried out at the NDS facility by TF81 personnel.
- Once the road-side processing was completed, three of five men brought out of the mosque were transported to an NDS facility in Kabul, known as NDS 90. Some TF81 personnel were involved in this, although most did not go to the facility but returned to base camp. Miraj was placed in a TF81 vehicle by TF81 personnel and was transported to the NDS facility where he and the other men were to be detained initially. It is possible that an Afghan national associated with TF81 accompanied the TF81 personnel in the vehicle but, on the evidence we heard, no one from the NDS or the ANP did. Miraj was seated in the middle of the back seat; it is probable that he was wearing plasti-cuffs, a blindfold and ear muffs. We understand that it was unusual for a person detained in a partnered operation to be taken to a detention facility by TF81 personnel in a TF81 vehicle—usually, detainees were transported in CRU or NDS vehicles.
- Upon arriving at the NDS facility, the men were escorted into a large lit-up room. TF81 personnel searched them again and a TF81 photographer took photographs of items seized and of the three men. As noted, the biometric processing was undertaken by TF81 personnel at this point, albeit, we were told, under the “supervision” of NDS personnel. As we understand it, TF81 personnel took the equipment for taking the biometric data on the operation and only they could operate it; further, the information gathered was for TF81/ISAF’s purposes; accordingly, it is not clear to us in what sense the NDS supervised (as opposed to observed) the process. A TF81 doctor assessed the men, albeit based on general appearance rather than a detailed medical examination. We heard evidence that at some stage, most likely as the detainees were being taken from the vehicles at the NDS facility, an NDS member spoke to one of the arrested men in Pashtun, which caused him to become visibly scared and very compliant.
- During this process, NDS staff came out and spoke to Afghan prosecutors who were with the group. It appears the handover process occurred relatively quickly. It is not clear what discussions there were, nor is it clear from the documents whether New Zealand had any way of tracking where Miraj was in the Afghan custodial system after he had been delivered to it.
- As we discuss in more detail later in this chapter, following the publication of Hit & Run in March 2017, the Chief of Defence Force ordered the military police to conduct an investigation into the circumstances of Miraj’s arrest and detention. A number of those interviewed described the practice of either hooding or blindfolding detainees, as well as the use of ear muffs and plasti-cuffs. Interviewees differed as to whether detainees’ hands were generally tied in the front or behind. While the Ground Force Commander for Operation Yamaha could not recall who placed Miraj in plasti-cuffs or whether he was hooded or blindfolded, he explained that the general use of restraints and blindfolds was to ensure that detainees were not able to obtain information that could later be used against New Zealand interests. In his interview with Mr Jon Stephenson, Miraj said that a cover had been placed over his head. In submissions, the Crown Agencies said TF81 used blindfolds, not hoods. Regardless of the specific method, it is clear that Miraj’s eyes were covered for the journey to the NDS facility.
- The operation at the mosque was relatively short, taking something between 30 minutes and one hour (from the time of entry to the completion of the tactical site exploitation). NZDF records show that the time from TF81’s initial deployment to the staging area at the airport to the time Miraj was located and detained was just over an hour and a half. The overall operation, from TF81’s initial deployment to withdrawal from the mosque site, was two and a half hours.
Was Qari Miraj assaulted by TF81 personnel?
- In the course of this operation no shots were fired and there were no civilian casualties. However, Hit & Run alleges that after Miraj had been placed in the vehicle and was waiting to be driven off, a member of TF81 reached into the back seat and started to shake him violently and then punched him repeatedly in the ribs and stomach and sometimes in the face. The book says that other TF81 members saw what was happening but did not intervene and that the attack was motivated by anger and a desire for vengeance.
- Following the publication of Hit & Run in 2017, Crown Law provided advice to the Attorney- General on the investigation options in relation to the allegations of mistreatment of Miraj. It is unnecessary to go into the details of that advice. It is sufficient to say that before charges could be laid, NZDF commanding officers had to determine that they considered the allegations to be “well founded”. To assist with that determination, the New Zealand Military Police began an investigation in May 2017. The Inquiry has had access to all the evidence collected during that investigation. All but one of the New Zealand soldiers present at the scene were interviewed and, according to the investigation, no evidence of mistreatment was found.
- In a briefing to the Minister of Defence dated 11 October 2017, the Chief of Defence Force recommended that the Minister note that the allegation of mistreatment of Miraj by TF81 personnel had been investigated and that no evidence was discovered that supported the allegations. The briefing stated that Miraj was detained; TF81 personnel searched him outside the building in which he was detained; after transportation to the NDS facility by TF81 he was again searched by TF81 personnel, photographed, and examined by TF81 medical personnel. According to this briefing, “[n]either the photos nor the medical examination revealed any signs of ill-treatment”.
- In light of its Terms of Reference, the Inquiry has itself conducted a further investigation into this issue, including by interviewing witnesses who were present and analysing contemporaneous documentary material. The Inquiry was also given access to transcripts of three interviews which Mr Stephenson conducted with Miraj in 2017, after the publication of Hit & Run. The interviews were conducted over the telephone and through an interpreter. During the interviews, Miraj acknowledged that he was an insurgent and said he had led the attack on the NZPRT patrol on 3 August 2010 in which Lieutenant O’Donnell was killed. In relation to his capture, Miraj said that after he was brought out of the mosque, he was hooded and left standing outside in the cold (the operation took place in the middle of winter). He claimed that he was punched by foreign forces, apparently before he was transported to the NDS facility.
- The Inquiry faced some difficulty in investigating this claim. First, as a result of the passage of time, personnel involved in the operation had difficulty recalling the detail of the operation. This meant, for example, that the Inquiry’s endeavours to ascertain precisely which personnel were in which vehicles was only partly successful.
- Second, most of those who gave evidence said that because the alleyway outside the mosque was narrow and dark and the vehicles were lined up there, there was little room and it was difficult to see what was happening other than in their immediate vicinity. Those in positions of responsibility said that they had not seen any indication of inappropriate behaviour during the operation. However, one said he had heard some gossip that Miraj had been punched—that some of the team were joking about it back at base. This soldier gave evidence which suggests that he spoke to the relevant person (person X) afterwards and person X said he had given Miraj a “bit of a touch up”. There was some suggestion that person X was known for making things up. In a separate conversation with another team member, the soldier joked that “it was probably [person X] that got punched (rather than the other way around).” It appears the soldier took no further action.
- Despite the difficulties in investigating aspects of the allegations, however, the Inquiry has, after careful consideration, been able reach conclusions based on the evidence it heard. The Inquiry received credible evidence from several sources (other than Miraj) to the effect that Miraj was punched, either once or several times, in the ribs or stomach as he was being placed into one of the three TF81 vehicles that went to the NDS facility. According to one witness, he did not complain too much because he was a “tough guy”. We also heard evidence from a number of witnesses about a persistent rumour that emerged soon after the operation that a particular TF81 member had punched Miraj. Some witnesses said that Miraj was “displaying a bit of attitude” when being placed into the vehicle, but the preponderance of the evidence was that he acted throughout in a low-key and compliant way.
- The medical officer who briefly examined Miraj could not recall speaking to him or the other men but did recall that Miraj was not displaying any obvious signs of injury or discomfort. The TF81 photographer described Miraj as quite relaxed. He said that Miraj “sort of found it humorous at times”.
- Finally, we note that person X, the individual who was alleged to have punched Miraj, was identified by witnesses and we put the allegations to him. He denied them vigorously.
- In the result, we are satisfied on the basis of the evidence that Miraj could only have been assaulted as he was being placed into the vehicle and was not assaulted while seated in the vehicle prior to departure to the NDS facility. On balance, we consider that Miraj was struck, either once or several times, in the ribs or stomach as he was being placed into the vehicle and while he was restrained and blindfolded. There is some difference, therefore, between the assault described to us and that described in Hit & Run, which is a more serious assault.
- It may not have been apparent to other TF81 personnel (that is, those who were not present when Miraj was being placed into the vehicle) that he had been assaulted given the confined space in the alleyway, the number of people captured and the need to deal with them all. We understand that Miraj did not complain at the time to TF81 personnel or to the Afghan national who was present. He also showed no obvious ill effects when he was observed by the TF81 doctor and photographed immediately before being handed over to the NDS at the NDS facility.
- There should have been some follow-up of this incident at the time, even though it was not a serious assault. The use of some force against a detainee may be justified in some situations, for example, to ensure a recalcitrant detainee enters a vehicle. But absent that, any assault of a detainee, whatever its severity, is unacceptable, especially when the detainee is restrained and blindfolded and poses no threat to the troops concerned. In addition to the requirements of operational policies and codes of conduct, there is a legal obligation to treat detainees appropriately, as we explained in chapter 10. In particular, Common Article 3 to the Geneva Conventions, which applies in non-international armed conflicts, requires that detainees be treated humanely and provides examples of prohibited acts. It prohibits a range of conduct and is not confined to the most serious forms of violence.
- Prompt, thorough and impartial examination of detainee mistreatment (whether alleged or suspected) is an important part of fulfilling applicable legal obligations. It is also important from a practical perspective: a person’s recollection of events is more likely to be fresh; people are more likely to give detailed descriptions of what they saw or heard; and there is a greater chance that other evidence or information can be gathered to inform an appropriate decision on next steps. Finally, prompt action signals the seriousness of detainee treatment issues and the expected standards of conduct and discipline in military settings.
What happened to Qari Miraj after he was transported to NDS?
- Hit & Run alleges that Miraj was tortured in Afghan custody and made a confession, from which NZDF benefitted. This is against the background of NDS’s reputation with regard to the treatment of detainees, particularly at certain facilities and in obtaining confessions, as we will discuss later in this chapter.
- The Inquiry examined why on this occasion the target of the operation was taken to an NDS facility. The evidence on this from those on the ground was somewhat vague. One explanation, for example, was that TF81 personnel did not know specifically where Miraj was to be taken, while others described that they learnt where they were going part way through the operation. However, we were advised by senior officers that it was TF81 who largely decided which Afghan force element would be responsible for the arrest and detention of particular suspects. On occasion, those arrested and detained by the CRU had been released within a day or two of their capture, in circumstances where their release did not appear to be justified. Obviously, with high value targets, such an outcome was thought to be undesirable. The IGIS’s report indicates that the deployed NZSIS staff facilitated the involvement of the NDS on this occasion.
- The Inquiry questioned the officers responsible for planning and carrying out the operation. They do not appear to have given much consideration to the risks of torture in the NDS facility to which Miraj was taken. They advised that they had information to suggest the particular facility was safe. Hon Dr Wayne Mapp gave evidence that some of the contemporaneous documentation identified this facility as ISAF’s “facility of choice” for operations of this nature. That said, there was general awareness among some TF81 personnel that the NDS had a reputation for mistreatment of detainees.
- In his interviews with Mr Stephenson, Miraj said he was tortured by the NDS when he was in detention—he was deprived of sleep, left cold in the middle of winter, beaten and given electric shocks. He said that NDS personnel were trying to make him confess to things he did not do (although he acknowledged that he was an insurgent and had been involved in particular actions against coalition and Afghan forces). Maulawi Neimatullah was also interviewed by Mr Stephenson about six months before Miraj’s first interview. He said that Miraj had been tortured while in prison, referring to sleep deprivation and electric shocks, and said that “they” had paid a substantial bribe to obtain his release (Miraj said he was released after being imprisoned for three years and eight months).
- The Inquiry has seen other independent evidence that supports Miraj’s claim that he was tortured. On 26 January 2011, following ongoing liaison with the NDS on the progress of their investigation into Miraj, the NZSIS formally received a copy of Miraj’s confession. The NZSIS disseminated the confession to a number of New Zealand agencies in a 31 January 2011 intelligence report. Later, on 25 February 2011, officials in New Zealand learnt of allegations that Miraj had been tortured. An assessment was made by Headquarters Joint Forces New Zealand that while there was probably an element of truth to the allegations, they were just as likely to have been deliberately contrived to influence the local population. In early March 2011, officials outside NZDF filed the information in a general oversight file for possible later legal review. We note that after assessing the report of the allegations and hearing evidence as to its context the IGIS was satisfied that “on its face there was good reason to assess it may well be true”. The allegations do not appear to have been brought to the attention of ministers.
- Like the IGIS, we consider that officials should have treated the allegations that Miraj had been tortured in NDS custody as being credible and requiring further enquiry. In assessing whether there was sufficient information to indicate that the allegations might well have been true, the specific circumstances relating to Miraj’s case had to be considered, as did the general context in which the allegations were made.
- In terms of case-specific circumstances, the following are relevant: the timing of the alleged conduct; the methods allegedly used; the circumstances, tenor and content of the later reporting and the likelihood of fabrication in that context; and observations made by New Zealand officers on the likely veracity or otherwise of the information. This case-specific evidence must be critically assessed against the most reputable international (United Nations and other) reports on the nature of treatment in NDS custody over the relevant period, as well as the wider international literature and guidance, which provides clear and useful information on the methods and patterns of torture in law enforcement and counter-terrorism or other high-security contexts. Taking these factors into account, we consider that there was strong evidence that Miraj was tortured, and that when they learnt of the allegations, New Zealand authorities should have made further inquiries and drawn the matter to ministers’ attention.
- Finally in this context, we note that when in late 2010 NZDF requested that the Government extend the NZSAS deployment to Afghanistan by 12 months, it addressed the question of detainee policy in the paper to the Prime Minister and the Ministers of Finance, Foreign Affairs and Defence (who were Ministers with Power to Act). The paper confirmed that NZDF would continue to apply the existing policy on detainees, which was summarised in an annexure to the paper. The detainee policy annexure contained the following statement:
In the event of the NZSAS becoming aware of any evidence of detainee mistreatment by the ANSF [Afghan National Security Forces], the NZSAS will be required to report in the first instance to CDF immediately and the matter raised with GIRoA [Government of the Islamic Republic of Afghanistan] as appropriate. If effective measures are not taken to rectify the problem this may require a cessation, on a temporary or permanent basis, of operations with the ANSF.
While, as we noted at paragraph , Headquarters Joint Forces New Zealand was aware of the allegations, we have seen nothing to indicate that the Chief of Defence Force was advised of them or that they were raised with the Afghan Government.
- The facts relating to the operation in which Miraj was captured are relatively simple and straight forward. The difficulties that arise revolve principally around the risks of torture to which he was exposed as a result of the agency and facility to which he was transferred. We return to this later in this chapter.
Developments in relation to detention policy following Operation Yamaha
- Following Qari Miraj’s capture in mid-January 2011, the issue of detainee treatment in Afghanistan, and the role of the NZSAS in detention operations, continued to occupy attention. Given that our Terms of Reference focus on the policy and its application to Qari Miraj, we will not canvass all subsequent events in detail. Rather, we mention aspects that we consider to be relevant to the issues we are considering, our principal focus being on the United Nations Assistance Mission in Afghanistan (UNAMA) report, released in October 2011.
- These subsequent developments are relevant to the issue of whether the transport and/or transfer of Miraj to the NDS in January 2011 was “proper” for several reasons:
- First, the UNAMA report dealt with detention conditions in Afghan facilities from October 2010 onwards, including the period when Miraj was captured and taken into detention. The report identified significant detainee mistreatment issues at some NDS detention facilities, including NDS 90, over this period.
- Second, the report’s findings had a significant impact and led to changes to ISAF’s approach to detentions on partnered operations.
- Finally, despite these developments, and the fact that NZDF and New Zealand agencies had earlier learnt of the allegations that Miraj had been tortured in NDS custody, New Zealand made no attempt to raise the matter with Afghan authorities or to investigate it further.
- In late April 2011, in response to parliamentary questions from Keith Locke MP about how many times the NZSAS had been “in the vicinity” when Afghan prisoners had been taken in joint operations, Dr Mapp advised that 35 individuals had been taken.
- In May, there was further media publicity about the issue. Dr Mapp’s concerns remained and he asked NZDF to provide a further briefing. On 31 August 2011, the Chief of Defence Force advised that the model of the CRU being responsible for effecting arrests was essential for “cultural, operational and developmental reasons, as well as legal ones”. No further elaboration was given on the legal reasons referred to. The tenor of the briefing also suggested a growing concern on the part of NZDF about the reputational impact of this topic. The briefing emphasised that NZDF personnel “do a vital job in a difficult and dangerous environment, often at risk to their own lives.” In the Chief of Defence Force’s view, allegations “that by so-doing they may be complicit in one of the most serious crimes are potentially damaging to the morale and mana of the members of the NZDF operating in this demanding theatre if not adequately addressed”. He advised that a small number of those detained by the CRU could be transferred to NDS, but reiterated earlier advice that NZDF was not responsible for transfer decisions; cited ISAF SOP 362 (which did not apply to partnered operations); and stated that NDS was the “detainee arrangement of choice” and that ISAF directed troops to make use of their facilities. As for the future, he advised the Minister that most of New Zealand’s future operations would require some form of collaboration.
- In September 2011, ISAF revised its approach to persons detained in partnered operations (which reflected the “two categories of detainee” approach) in anticipation of UNAMA’s report. The forthcoming report focused on the treatment of detainees regardless of how they came to be in Afghan custody (that is, whether they were captured in a wholly ISAF operation or in a partnered operation). Due to the nature of the material, we are unable to describe ISAF’s revised approach in much detail, although a couple of relevant points are worth noting. First, the concept of detention was widened to include partnered operations. Second, in addition to the steps ISAF would undertake (for example, training, inspections and a temporary suspension of transfers to certain facilities, including NDS 90 where Miraj had been taken), there was an expectation that states would also review their own policies in light of the changes and implement a range of additional measures to protect detainees captured on partnered operations from abuse, both before and after transfer.
- On 10 October 2011, UNAMA released its report. By way of background, a year earlier in October 2010, in response to complaints the organisation had received from communities within Afghanistan and others, UNAMA instigated a detention observation programme focusing on detainees held for offences related to the armed conflict in Afghanistan. This involved visiting facilities and interviewing a number of conflict-related detainees between October 2010 and August 2011. For this report, UNAMA explained that a detainee was considered to be “conflict-related” if they were charged with crimes against the state under Afghan criminal law, or with relevant terrorism offences. UNAMA interviewed detainees who had been captured in operations where ISAF forces acted alone, as well as operations in which ISAF forces acted jointly with local Afghan forces. The report provided the results of this observation programme. UNAMA had been granted access to all NDS detention facilities except two, one of which was NDS 90 in Kabul (where Miraj was taken). However, by interviewing detainees who had spent time at these two facilities, UNAMA was able to gather information about conditions in them.
- Of the 28 individuals interviewed about NDS 90, 26 reported torture, which included beating, suspension, twisting or wrenching of genitals and, in some cases, the use of electric shocks. All abuse was reported as having occurred during the interrogation process. UNAMA interviewed 126 detainees about treatment in 17 further facilities not covered in the UNAMA report, including NDS 17 in Kabul. Of these, 25 per cent alleged they had been tortured. UNAMA required further time to establish the credibility of these allegations and, at the time of the release of its report, continued to investigate the accounts provided. In response to the suggestion by some governments and others that detainees make false allegations “as a form of anti-Government propaganda”, UNAMA addressed how it ruled this out when assessing the veracity the accounts provided during interviews with detainees.
- In relation to the matters that were verified through the observation programme, UNAMA concluded:
UNAMA’s detention observation found compelling evidence that 125 detainees (46 percent) of the 273 detainees interviewed who had been in NDS detention experienced interrogation techniques at the hands of NDS officials that constituted torture, and that torture is practiced systematically in a number of NDS detention facilities throughout Afghanistan. Nearly all detainees tortured by NDS officials reported the abuse took place during interrogations and was aimed at obtaining a confession or information. In almost every case, NDS officials stopped the use of torture once detainees confessed to the crime of which they were accused or provided the requested information.
- This was despite the fact that the Afghan Constitution explicitly prohibited the use of torture “even for discovering the truth from another individual who is under investigation, arrest, detention or has been convicted to be punished”. The use of torture was also a criminal offence under the Afghan Penal Code. UNAMA found that five facilities, which included NDS 90, systematically tortured detainees for the purpose of obtaining a confession. UNAMA investigators interviewed 89 detainees who reported the involvement of international forces (either alone or partnering with Afghan forces) in their capture and transfer to the custody of either NDS or the Afghan National Police (ANP). In relation to 19 of these, UNAMA found compelling evidence of torture while in NDS or ANP custody.
- Among other things, UNAMA recommended that troop-contributing states: suspend transfers to NDS and ANP facilities where credible allegations of torture existed; review monitoring practices at each NDS facility where detainees are transferred and revise as necessary to ensure detainees are not transferred to a risk of torture; review transfer policies to ensure adequate safeguards; use their joint operations, funding arrangements, the transition process, intelligence liaison relationships and other means to stop the use of torture; and build capacity through mentoring and training of NDS and ANP on the legal and human rights of detainees and on effective interrogation methods and forensics.
- A few weeks before the release of the report, the then Director of Defence Legal Services, Brigadier Kevin Riordan, provided advice on a draft which had been provided to NZDF. He noted that the allegations in the report were regarded as credible, but emphasised that NDS 17 had only a few allegations in respect of it and that partnering arrangements with CRU would not change. Others who were considering UNAMA’s report had noted that, although not officially sanctioned or directed, the report appeared to have uncovered torture on a large scale. Officials appear to have paid little attention to NDS 90. While NDS 17 may have been front-of-mind given the assessment in Evans, NDS 90 equally warranted attention: UNAMA had addressed NDS 90 in its report and ISAF had taken specific steps as a result. NDS 90 was located in Kabul and was one of the facilities where detainees captured in TF81/Afghan operations were transferred, including Miraj.
- Dr Mapp sought advice from officials. In particular, he noted that the report seemed to require specific action from New Zealand and asked his Press Secretary to find out exactly what New Zealand was doing. NZDF advised that no-one captured by New Zealand was in NDS custody but did not mention partnered operations. NZDF said that its policies on transferring detainees had been reviewed and were considered robust and appropriate to deal with the concerns in the UNAMA report. No further information was provided about the basis for this assessment.
- According to NZDF, the “headline point” was that most of UNAMA’s recommendations were to increase training and mentoring. NZDF emphasised that UNAMA had not made a complicity finding and, in fact, recommended more mentoring and training. It repeated earlier advice that New Zealand had no resources or legal power to monitor detainees and stated that, to the best of its ability to ensure, NZDF was confident no CRU detainee had been mistreated or handed over to prohibited facilities. Among other things, the advice was silent on those transferred directly to NDS 90 (such as Qari Miraj), and on those the CRU transferred to the NDS.
- Documents available to the Inquiry show that around this time, Brigadier Riordan recommended that New Zealand approach the Afghan Government to express concern about what was revealed in the UNAMA report and to seek assurances not only that people captured on partnered operations in the future would not be tortured but also that those captured in the past with New Zealand assistance had not been tortured. No action was taken, however, apparently on the basis of the Solicitor-General’s advice.
- On 11 October 2011, agencies also prepared written advice for the Minister of Defence and Prime Minister that ISAF was managing the situation and New Zealand had only captured one detainee. Separately, NZDF also advised, as it had done in an earlier briefing to the Minister of Defence, that only a small number of persons arrested by the CRU are transferred to the NDS in Kabul. NZDF figures from that period provide a different picture. Of 160 total operations, 74 resulted in detention of 193 individuals. In 58 of those operations, 150 individuals were taken into Ministry of the Interior custody (it is not clear where they were sent thereafter). On 15 operations, 42 individuals were taken into NDS custody. By these figures, just over one fifth (21.8 per cent) of all detainees taken in partnered operations were sent to an NDS facility. It is also possible that some of the 153 persons transferred to Ministry of the Interior custody may have been later transferred to the NDS.
- Later in October, the Chief of Defence Force provided further briefings to the Minister of Defence. These briefings emphasised that torture was not systematic in all facilities; it was not a de facto policy; and NDS had cooperated with UNAMA. They also advised that, since the completion of the UNAMA report, no one arrested during CRU operations had been taken to prohibited facilities. This advice did not address whether persons transferred before this date, such as Miraj, had been transferred to prohibited facilities nor did it mention that UNAMA had been prevented from accessing NDS 90.
- The Minister of Defence publicly released the NZDF briefings to him of 31 August and 20 October 2011. Before this, officials identified potential questions and apparent inconsistencies between the advice and what was in the public domain on NDS. One official, for example, observed that the fact of the Evans decision implied New Zealand “either know there is a risk of torture or it is reasonable to assume we know but the CRU we mentor still continues to use the facility.” After publication, officials discussed with ministers the possibility of seeking assurances from Afghanistan about how persons captured during partnered operations had been treated. It does not appear that officials recommended monitoring but the instruction to seek assurances was given, although it is not clear if these assurances were in fact sought or obtained.
- On 21 October 2011, the Chief of Defence Force issued an all-staff email, in which he advised that NZDF complied with United Nations and international standards for detainee transfers. He emphasised that torture was not institutionalised in all facilities and UNAMA had advised that countries “must continue” to partner with local authorities. He said that he regarded it as his duty to protect staff from being implicated in any breach simply through carrying out their roles, which they did in a difficult and dangerous environment. While the Chief of Defence Force no doubt wanted to maintain staff morale, some of the key messages in this email were inaccurate and others ought to have been more carefully considered.
- Officials briefed the Minister of Defence on what ISAF was doing in response to the UNAMA report. These briefings, however, did not inform relevant ministers of the specific expectations of units involved in partnered operations, nor that persons detained in previous partnered operations (such as Miraj) had been taken to facilities to which transfers had been suspended as a result of the UNAMA report. It does not appear that the relevant ISAF material was provided to the Minister.
- In late November 2011, shortly after the release of the UNAMA report and ISAF’s updated approach to partnered operations, New Zealand was informed of fresh allegations of torture made by detainees in NDS facilities. New Zealand also learnt that NDS 17 (the facility assessed in Evans) and NDS 90 (the facility to which Qari Miraj had been transferred) were located on the same compound and that detainees were sometimes transferred between facilities during the investigation phase. This information was significant in terms of assessing the risk of torture to persons transferred into Afghan custody and NZDF staff raised concerns internally with NZDF in Wellington about the legal risks associated with continuing transfers to NDS facilities. Very shortly after the initial reporting, however, New Zealand was advised that the allegations as to treatment could not be substantiated and there was no evidence of systematic transfer of detainees between facilities. New Zealand officials appear to have accepted this assessment without further meaningful follow-up. It also appears that the Minister of Defence was never briefed about this—that is, the fact that serious allegations had been made, the initial feedback on the allegations and ISAF’s final assessment of the evidence, and on the possibility of transfers between facilities.
- We draw attention to two points arising in this sequence of events. The first is that UNAMA commenced its investigation in October 2010, in response to complaints it had received about the situation in Afghan detention facilities. UNAMA came to similar conclusions as the Court in Evans had, namely that there was a significant and continuing problem with the mistreatment of detainees in some NDS facilities. In our view, the Court’s judgment in Evans should have alerted New Zealand to the fact that there was a significant and continuing problem with what was occurring in Afghan detention facilities that required an active response. This, after all, was the effect of the legal advice given by Brigadier Riordan and the Solicitor-General following the decision.
- We acknowledge that Dr Mapp had discussions with Afghan counterparts about detention issues during his visit to Afghanistan in late August 2010, and was encouraged by what he was told. We also acknowledge that he participated in international discussions about the issue and was generally concerned about it. As well, we accept that New Zealand officials engaged with the officials of other ISAF partners on detention issues, and to some extent with Afghan officials. But little effective action seems to have been taken after the receipt of the legal advice, which made it plain that more active and effective engagement with the issue was required. This is despite the fact that the Chief of Defence Force raised the matter specifically with the Minister in his briefing note of 16 September 2010.
- The second point is that once UNAMA’s findings became available, ISAF did reconsider and revise its position in relation to detentions on partnered operations. New Zealand officials also took some steps, but they were modest and, in some cases, would not have provided meaningful or effective protection to individuals from real or immediate risk of torture. Our impression from reading the documents is that officials were reluctant to depart from the view they had expressed earlier in 2010 that there had been “substantial improvement” in conditions of detention in Afghan facilities.
- We now give our assessment of New Zealand’s detention policy in relation to persons detained during partnered operations and its application to the capture and detention of Qari Miraj. We do so under the following headings:
- A negotiated arrangement?
- New Zealand’s detention policy for partnered operations examined.
- How should Qari Miraj’s capture and subsequent detention be analysed?
We also address New Zealand’s failure to take steps after learning of the allegations concerning Miraj’s torture.
A negotiated arrangement?
- As the Crown Agencies acknowledged in submissions, New Zealand could, had it so wished, have attempted to enter into an arrangement with Afghanistan in relation to people arrested by Afghan personnel in partnered operations, as it had done in respect of those detained by NZDF directly and transferred to Afghan authorities. The arrangement that New Zealand entered into with Afghanistan in the latter context was an acknowledgment that the persons captured were under the “effective control” of New Zealand forces, so that New Zealand had “jurisdiction” over them, from which certain legal obligations flowed. The arrangement placed constraints on the way Afghanistan could deal with people transferred to its custody by NZDF personnel, and provided monitoring and similar mechanisms designed to give New Zealand the ability to check that Afghanistan was meeting its commitments. However, that curtailment was not an unjustified interference with Afghanistan’s sovereignty because it was something to which Afghanistan agreed following negotiations. In other words, the “curtailment” resulted from an exercise of sovereignty.
- There was no reason of principle that prevented New Zealand from pursuing a similar arrangement with Afghanistan in relation to those who were arrested by Afghan personnel during partnered operations, as Brigadier Riordan noted in his opinion after the Evans decision was delivered. Such an arrangement could have been presented to Afghanistan as a requirement for New Zealand’s participation in partnered operations, just as New Zealand had required a written arrangement in relation to direct detentions and subsequent transfers to Afghan custody. It would then have been up to Afghanistan, in the exercise of its sovereignty, to have decided whether it was prepared to enter into such an arrangement. Afghanistan would, of course, have been free to decide that it did not wish to enter into any such arrangement. If it declined to do so, New Zealand would then have had to decide about the legal and other implications of that, including whether, or how, to proceed with the partnering arrangement with the CRU. The Crown Agencies submitted that this had the potential to undermine New Zealand’s engagement in Afghanistan in furtherance of the Security Council resolutions, although Security Council resolutions do not, of course, relieve states from meeting their international obligations.
- As is reflected in the Solicitor-General’s advice, any arrangement with Afghanistan in relation to persons detained on partnered operations need not have contained the same elements as the arrangement in relation to persons detained directly by NZDF, so long as the ultimate purpose of the policy (effective protection of individuals from the risk of torture) could be maintained. Besides training and mentoring, an arrangement could have involved a conditional assistance policy, in which assistance was varied, amended, suspended or terminated if there were indicators of abuse in relevant Afghan detention facilities; some form of monitoring of conditions in particular detention facilities; and so on. In short, an arrangement involving something less onerous than monitoring of each individual detainee might well have been appropriate, at least for most detainees, as long as it was, overall, effective.
- While the Crown Agencies acknowledged that New Zealand could have attempted to negotiate some sort of arrangement with Afghanistan in relation to the treatment of detainees arrested in partnered operations, it said that doing so would have proved difficult. That may be correct; but having already negotiated an arrangement with Afghanistan in respect of one category of detainee, it may have been possible to negotiate another arrangement more readily. An ambitiously crafted arrangement might also have included provision for New Zealand to provide more hands-on training and capacity-building of Afghan officials to enhance human rights compliance. Given that NZDF was in theatre for the purpose of giving effect to the UN Security Council mandate, an arrangement which directly addressed detention issues may, if implemented well, have been beneficial to the longer-term objectives of that mandate. In any event, there was no attempt to negotiate any form of arrangement—indeed, the option was specifically rejected.
- The contemporaneous documents indicate that the option was rejected for two main reasons. The first was that New Zealand agencies considered that there was no legal obligation to enter into such an arrangement; the second was a concern about the resources required.
- Putting to one side for the moment the question whether there was a legal obligation, we reiterate that both Brigadier Riordan and the Solicitor-General considered that there may well be at least a moral obligation on New Zealand to take some steps in relation to those arrested by Afghan authorities on partnered operations, although they did not agree on precisely what those steps should be. Brigadier Riordan had in mind a formal arrangement with Afghanistan, with the ability to monitor the treatment of individual detainees and so on. The Solicitor-General counselled against setting up a monitoring programme for individual detainees, but agreed that steps were required, such as seeking formal and operational assurances, gathering information about circumstances within Afghan detention facilities and, if it became apparent that torture was occurring, taking steps in response. But the point for present purposes is that both recognised that there was a moral dimension to this issue going beyond the strictly legal one. Although the contemporaneous documents indicate that other agencies such as MFAT were prepared to acknowledge this, the issue of resources seems to have been a factor early on in the deployment and was a significant factor throughout.
- We are troubled that the option of seeking some form of arrangement with Afghanistan about the treatment of detainees captured on partnered operations was rejected. New Zealand deployed forces to Afghanistan in 2009 in the knowledge that, despite the heavy commitment of military and humanitarian resources to Afghanistan since September 2001, torture and other mistreatment still occurred in Afghan institutions. As we noted earlier, officials were aware of the international reports concerning conditions in Afghan facilities. New Zealand was careful to make appropriate arrangements in relation to persons it detained directly, yet took the view in its detention policy that an arrest by Afghan authorities on partnered operations effectively insulated it from significant obligations in relation to torture or other mistreatment of detainees.
- We consider that New Zealand had at least a strong moral obligation to provide its support to Afghanistan in a way that ensured, to the extent possible, that its actions did not contribute to people being tortured or mistreated in detention. We see this as reflecting the fundamental nature of the prohibition on torture under international law and its associated obligations, and New Zealand’s long-standing commitment to international human rights norms. This gains additional support from the legal obligation which Sir Kenneth Keith highlighted in Common Article 1 of the Geneva Conventions of “ensuring respect” “in all circumstances” for the obligations towards detainees flowing from Common Article 3, to which we return below, as well as from the other sources of preventive obligations discussed in chapter 10.
- It must be remembered that Afghanistan’s criminal justice system, inherited from the Soviet regime, was a confession-based system. As both the judgment in Evans and the later UNAMA report indicated, most torture occurred within a short time of suspects being detained and was aimed at obtaining confessions. “High value” detainees from whom valuable intelligence might be obtained were likely to face greater risks than ordinary criminal suspects. Consequently, the risk of torture existed in both law enforcement and armed conflict settings. Given that an important role of NZSAS personnel in Afghanistan was to mentor the CRU, a specialist police unit, as it performed law enforcement functions, we consider New Zealand could and should have recognised the reality of the risk of torture as a result of operations in which its forces were involved by making appropriate arrangements with Afghanistan in relation to people arrested by Afghan authorities on partnered operations.
- In their submissions, the Crown Agencies noted that numerous other countries which contributed forces to ISAF adopted the same policy as New Zealand towards detainees on partnered operations, as did ISAF itself. The Crown Agencies submitted that this constituted evidence of states’ practice, which is relevant to determining the scope of obligations under customary international law and the interpretation of obligations under international treaty law.
- However, we do not know how many states engaged in partnering arrangements with Afghan forces, what the nature of those partnering arrangements were or how they were implemented in practice; for example, we do not know what the precise role of other states’ forces on partnered operations was. We are also unable to review the detention arrangements of all countries which contributed to ISAF. Consequently, we are not in a position to fully assess the basis of this claim. But even if (a) it were true and (b) we assume that arrangements in a particular non-international armed conflict could be treated as decisive in terms of states’ practice, questions about New Zealand’s position would remain. New Zealand has tended to view itself as being at the forefront in matters of international human rights and has in the past adopted values- and rights-based stances in advance of other comparable states. The positions adopted by other states should not necessarily determine the scope of the obligations that New Zealand sees as governing its conduct, or the way in which it discharges those obligations. This became particularly acute after the Solicitor-General and others raised the possibility that there were relevant legal obligations. That suggested that a precautionary approach should be adopted, one which afforded the greatest level of protection to individuals and which reflected the humanitarian principles underlying this area of law. In any event, it would, in our view, be unfortunate if New Zealand lost its commitment to independent thinking on important issues of human rights such as the prohibition on torture and the effective prevention of it.
- To conclude, we think it disappointing that New Zealand chose not to attempt to enter into an arrangement with Afghanistan about the treatment of persons detained by Afghan authorities on partnered operations given the information available to it about conditions in Afghan detention facilities, which was subsequently confirmed by Evans in 2010 and the UNAMA report in 2011. Throughout the Operation Wātea deployment, there was information available that, despite some progress, torture was still prevalent in Afghan detention facilities, particularly NDS facilities. This information, together with the information about the confession-based nature of the Afghan criminal justice system, should have led New Zealand to appreciate that there was a real risk that some of those arrested by Afghan personnel on partnered operations would be detained in facilities where they could well be subjected to torture. Even if not legally obliged to do so, we consider that New Zealand should have recognised and addressed this in some form of over-arching arrangement.
- It is clear from the contemporaneous documents that the resource implications of taking more active steps in relation to those detained on partnered operations were given significant weight. This feature was also raised by Counsel for the Crown Agencies in submissions, by NZDF and by Dr Mapp as a justification for New Zealand’s approach. We think it unattractive that considerations of that type should be given significant weight (or worse, be determinative) in a context such as this, especially given New Zealand’s oft-stated commitment to being a good international citizen. If there are strong obligations of this sort, surely New Zealand should either put itself in a position to meet them or not engage in the enterprise in the first place—all the more so, of course, when the obligations are legal ones.
New Zealand detention policy for partnered operations examined
- We now come to the detail of the policy. We begin by recording the obligations in relation to torture that the Crown Agencies acknowledged applied to New Zealand forces in partnered operations:
- First, the Crown Agencies accepted NZDF personnel were obliged to treat people humanely during interactions. They could not commit torture themselves or acquiesce in Afghan personnel torturing people. The Crown Agencies also accepted that the training and mentoring provided by NZDF personnel to Afghan personnel should encourage respect for International Humanitarian Law and International Human Rights Law.
- Second, NZDF personnel had to avoid becoming complicit in torture.
- Third, there could be factual situations where a person captured in a partnered operation came under New Zealand’s jurisdiction so that non-refoulement obligations could apply. This depended on the nature and extent of New Zealand’s involvement in particular operations. If in an operation NZSAS personnel were not in fact providing support to Afghan authorities in a law enforcement context but were acting for their own purposes and Afghan authorities were “merely co-opted”, it might be appropriate to classify the operation as an NZSAS operation with Afghan authorities effectively acting simply as agents of the NZSAS. In addition, the Crown Agencies accepted that there might be instances where, although the intention was for Afghan authorities to conduct an arrest, an operation could turn from a “law enforcement” to an “active hostilities” operation, during which the NZSAS detained a person under the authority provided by the relevant Security Council resolutions and International Humanitarian Law.
As will be apparent, then, the Crown Agencies did not accept that New Zealand had non-refoulement obligations in relation to detentions on partnered operations in the ordinary course, but did accept that the particular facts of operations matter, so that in some circumstances New Zealand might have such obligations.
- According to the Crown Agencies, the critical question for determining on which side of the line an operation falls is whether it “could properly be considered to have been conducted in order for the Afghan authorities to exercise their law enforcement function, and whether the participation of the NZSAS can properly be said to have been directed at supporting this”. Consequently, as long as the purpose of the involvement of the NZSAS in an operation was to provide support to the Afghan authorities to effect an arrest and their participation can be seen as supporting this, it would not matter that the operation was initiated, planned and substantially conducted by NZSAS personnel or that it could accurately be described as “New Zealand-led”; nor would it matter that the warrant issued and executed by Afghan authorities had been prepared by NZSAS personnel. The fact that:
- it was intended that Afghan officials would execute an arrest warrant; and
- Afghan officials did that;
would mean that New Zealand would bear no responsibility if mistreatment occurred in detention, even where that was a real risk at the time of capture: in effect, that was none of New Zealand’s business.
- As we see it, the position adopted by the Crown Agencies is based on the premise that there was a sharp distinction between armed conflict and law enforcement operations. However, as we noted in chapter 7, this distinction is problematic. As the Court in Evans recognised, detention operations had both armed conflict and law enforcement dimensions. The insurgency in Afghanistan presented challenges because of the ability of insurgents to merge into the general population and from the forms of warfare they engaged in—surprise guerrilla-style attacks, use of improvised explosive devices and so on. Counter-insurgency activities posed considerable threats to the forces involved in them. Equally, however, Afghan domestic authorities were interested in prosecuting insurgents, given that their activities almost inevitably involved the commission of crimes (for example, terrorism offences). So, the detention of an insurgent was likely to serve both armed conflict and law enforcement interests, albeit that armed conflict and law enforcement activities have different objectives. This is because armed conflict seeks to address immediate and future threats to safety, whilst law enforcement is concerned with punishing past infractions. Consequently, using a test which looks at operations based on their purpose, or the purpose of one of the forces involved, may not be particularly useful. There may be more than one purpose to an operation, and to a force’s participation in that operation. This is an important part of any factual or contextual assessment.
- We are troubled by the Crown Agencies’ analysis for another reason. We take Operation Burnham as an example. If the two targets, Neimatullah and Kalta, had been captured during the operation as described in chapter 4, we consider it could not sensibly have been argued that NZDF personnel did not have “effective control” over them, even if at some stage of the operation they were formally arrested by Afghan personnel pursuant to Afghan arrest warrants. We say this because NZDF personnel were responsible for all meaningful elements of Operation Burnham—obtaining the necessary intelligence; undertaking the necessary liaison with ISAF; obtaining the necessary air support; and planning, organising, leading and undertaking the operation. The operation was far beyond the operational experience, skills and capabilities and equipment levels of the CRU personnel involved. For security reasons, CRU personnel were not advised of the operation until a day or two before it occurred. The planning documents make two things clear—first, that this was an operation motivated by armed conflict considerations (in particular, the protection of the NZPRT) rather than law enforcement ones, and second, that NZDF personnel considered they were entering an area of Taliban influence and anticipated the likelihood of armed resistance.
- As we understand it, the Crown Agencies accepted that the planning for Operation Burnham proceeded on an armed conflict basis, and that if Neimatullah and Kalta had been captured during the operation as it in fact played out, New Zealand would have assumed jurisdiction over them for the purpose of the application of non-refoulement obligations. However, the Crown Agencies argued that if Neimatullah and Kalta had responded to the “soft knock” and surrendered, thus allowing the Afghan officials to move forward from the helicopter landing zone and execute the arrest warrants, the operation could properly be considered to be in support of Afghan authorities pursuing legitimate domestic law enforcement functions. On that scenario, New Zealand would not have assumed jurisdiction over them.
- We find this analysis difficult to accept. While we agree that facts matter, we do not accept that the fundamental nature of an operation can be recast or reconstructed in the way this argument suggests. On the hypothetical scenario just outlined, the operation would have been planned and carried out as a highly complex armed conflict operation. The only significant role played by Afghan personnel would have been to execute arrest warrants prepared by NZDF personnel. We find it difficult to accept that that feature then becomes the decisive factor in terms of the application of non-refoulement obligations. Rather, we consider that weight must be given to the substance of what the operation entailed, particularly given the fundamental importance of the obligations at issue.
- In summary, to determine whether a particular person was under the “effective control” of New Zealand or Afghan personnel, or possibly both, we consider that the substance of what occurred on a partnered operation must be considered. We accept that there would have been operations where NZDF personnel did not take a leading role in the “active part” of the operation (as Brigadier Riordan put it in his opinion), but simply provided mentoring support and technical or other assistance to CRU personnel as they performed law enforcement functions. In such operations, we accept that New Zealand personnel would not have had jurisdiction for the purposes of obligations in relation to non-refoulement. They would, in a real sense, simply be “mentoring” their Afghan partners. But that cannot be said of all partnered operations.
- Finally, before we leave the topic of the policy, we should note that, as discussed in chapter 10, New Zealand had relevant obligations besides those relating to non-refoulement.
- First, Sir Kenneth Keith drew attention to Common Article 1 to the Geneva Conventions, under which states must “ensure respect” for, among other things, Common Article 3, which prohibits the mistreatment of, and requires minimum standards of care for, detainees. Sir Kenneth said that states were required to do everything reasonably within their power to ensure that other states complied with these obligations. In concluding his expert evidence to the Inquiry, Sir Kenneth said:
In the present situation the particular characteristics of the provision of “partnering, including close support and technical support” or more generally the “provision of assistance” by the NZDF with the Afghan authorities may well be decisive in determining whether the NZDF is in breach of the duty to ensure respect, to the best of its ability, for the prohibition on torture in terms of Articles 1 and 3 or is complicit in torture under customary international law.
This indicates that the Common Article 1 obligation operates at an individual level, not simply a systemic one. As previously noted, the obligations under Common Articles 1 and 3 overlap the various other treaty and general international law obligations to prevent torture.
- Second, the Solicitor-General said that New Zealand had an obligation of non-complicity, which required it to ensure that it did not provide assistance to Afghan personnel where it was known, or ought to be known, that the assistance provided “sufficiently direct support” for acts of torture by Afghan authorities. In partnered operations, the Solicitor-General saw this obligation as arising at a systemic level, requiring New Zealand to take steps to deter torture by Afghan authorities, gather information about the practices of Afghan personnel and institutions and, if necessary, restrict or withdraw cooperation until matters were remedied (for example, by varying or suspending partnering arrangements).
- Third, the Solicitor-General said that it was possible that even where CRU operations were formally a matter of law enforcement, they might engage members of organised armed groups “at a sufficient level of intensity to engage the law of international armed conflict”. So, he said, it was prudent to anticipate that New Zealand forces might, in some operations, also become subject to International Humanitarian Law considerations. The Solicitor-General identified a relevant obligation as being the duty to “ensure respect” for Law of Armed Conflict standards, an obligation which he considered to be applicable in non-international armed conflicts such as Afghanistan. He noted that the ICRC had described the obligation as requiring states to exert influence, to the degree possible, to stop violations of International Humanitarian Law. He said this might include steps such as diplomatic pressure and withdrawal of co-operation or assistance.
- Against this background, we turn to consider the facts of Operation Yamaha.
How should Qari Miraj’s capture and subsequent detention be analysed?
- The Crown Agencies argued that as New Zealand did not have “effective control” over any area or facility where Qari Miraj was arrested or detained in January 2011, he could only be subject to New Zealand’s jurisdiction by way of personal jurisdiction. While the Crown Agencies acknowledged that there were decisions of the European Court of Human Rights which employed a wide concept of personal jurisdiction, they argued that the concept of personal jurisdiction contemplated by the Convention against Torture or by the ICCPR was not as wide as that applied in these cases.
- While we do not consider that this view necessarily reflects the international law or practice of torture prevention as it has developed, we do acknowledge that even the European authorities currently consider that something more than the “mere” use of force or the short-lived application of some control by a state agent acting outside the state’s territory is needed to bring the “detainee” within the jurisdiction of that state.
- In any event, once it is accepted, as New Zealand accepts, that a state operating outside its territorial jurisdiction may acquire obligations in relation to torture based on personal jurisdiction, it becomes necessary to identify when a state acquires such jurisdiction. This depends on an analysis of the facts—in this case, an analysis of the nature and details of Operation Yamaha.
- The Crown Agencies acknowledged that where TF81 personnel captured a person when acting on their own, they had personal jurisdiction over the person and New Zealand’s non-refoulement obligations were engaged. However, in relation to Operation Yamaha, they submitted that TF81’s role in it was properly described as supporting Afghan authorities to conduct an arrest in pursuit of domestic law enforcement objectives, even though the operation could fairly be characterised as a “New Zealand-led” or a “joint” operation. The Crown Agencies said that there were interactions with Afghan authorities in the months prior to the operation about a possible operation, which showed the “law enforcement” nature of the operation. Further, although TF81 could have detained Miraj under International Humanitarian Law, in fact he was arrested under Afghan law and the limited period that TF81 had physical custody of him was insufficient to bring him within New Zealand’s jurisdiction. They also argued that even if New Zealand’s non-refoulement obligations were engaged, New Zealand could reasonably have concluded that Miraj did not face a real risk of torture, given the information available to it. Finally, the Crown Agencies submitted that there was no intent or knowledge of the type required for a finding of complicity.
- We address the issues under two headings:
- Did New Zealand owe non-refoulement or similar preventive obligations to Miraj?
- If so, did New Zealand breach them?
- Before we do so, however, we should address an important preliminary point. As we noted in chapter 1, the Inquiry’s Terms of Reference echo the Inquiries Act 2013 by stating that the Inquiry “has no power to determine the civil, criminal, or disciplinary liability of any person” although it may “make findings of fault …”. The Crown Agencies submit therefore that the Inquiry may not state whether it considers New Zealand was in breach of a relevant obligation to Miraj, as this amounts to a finding as to the civil liability of New Zealand at international law.
- The Inquiry accepts that it has no power to determine the civil liability of any person and does not purport to do so. It does not accept, however, that it is not entitled to express its view about whether (or not) New Zealand owed any relevant obligations to Miraj and, if so, whether it met them.
- Under s 11(2) of the Inquiries Act, an inquiry may make findings of fault in the course of “performing its duties under this Act”. Clause 7.8 of the Terms of Reference requires the Inquiry to report on whether NZDF’s “transfer and/or transportation” of Miraj to the NDS in Kabul was “proper”, given (amongst other things) the decision in Evans. We cannot answer this question without reaching and expressing a view about whether New Zealand had international obligations toward Miraj and, if so, whether it met them. Any view we express on that is our own and is not a final determination of New Zealand’s liability one way or another. It will simply be a step in the reasoning that we must undertake so as to answer the question that has been put to us.
Did New Zealand owe non-refoulement or similar obligations to Miraj?
- Operation Yamaha was not, in our view, a straightforward law enforcement operation conducted by Afghan personnel with TF81 personnel simply “in the vicinity” or adopting an oversight or mentoring role in respect of their actions. TF81 was concerned to ensure Qari Miraj’s capture because he had played a key role in the 3 August 2010 attack in which Lieutenant O’Donnell died; as a result of that attack, he was a demonstrated threat to the NZPRT in Bamyan; intelligence after Operation Burnham indicated he was likely to attack again; the NZPRT Commander had contacted TF81 in September 2010 seeking their assistance in capturing him; and he was listed on the JPEL at New Zealand’s instigation and so was a legitimate ISAF target. Against this background, New Zealand had kept a close eye on Miraj’s movements to identify an opportunity to mount an operation against him. From TF81’s perspective, Operation Yamaha was undoubtedly important to New Zealand and had a significant “armed conflict” component to it. Indeed, given the NZPRT Commander’s concerns and the fact that planning for Operation Burnham had indicated other ISAF force elements had other priorities, removing Miraj as a threat was akin to a New Zealand national task. We note that NZDF told us that Operation Burnham was not a law enforcement operation or mentoring exercise for the CRU, but a national task approved by the Chief of Defence Force. Although Operation Yamaha was not an “out of area” operation, we see it in the same way as NZDF described Operation Burnham.
- Moreover, we consider that Afghan authorities were not involved in the planning or preparation for Operation Yamaha in a meaningful way. In this connection, we note that New Zealand personnel received the information leading to the operation and passed it on to TF81, who then obtained the necessary air asset support and other approvals and undertook the planning and organisational work for the operation. We note that TF81 did not perform this operation with the Afghan partner force they were mentoring (the CRU), but with the NDS in conjunction with the ANP. While it appears that TF81 had previously had some involvement with the NDS, it is not clear to us what relationship they had with the ANP. In any event, TF81 personnel were not working with the Afghan partners whom they had been training for the previous year or so, which no doubt affected the way the operation was carried out.
- In addition, post-capture processing of Miraj (and those arrested with him) was carried out by TF81 personnel, not by Afghan personnel. This included steps taken at the roadside outside the mosque and at the NDS facility in Kabul to which TF81 personnel transported Miraj. These steps included formally identifying Miraj, plasti-cuffing and blindfolding him, searching him, seizing his cell phone so that information could be taken from it, photographing and taking biometric data from him and having a doctor observe him, after which he was handed over to NDS personnel at the facility.
- Further, New Zealand had a strong interest in the fruits of the operation. As we noted earlier, the biometric processing equipment was controlled and operated by TF81 personnel, not Afghan personnel, and TF81 personnel took the biometric data for TF81/ISAF’s use. In addition, TF81 personnel retained and exploited the information obtained from Miraj’s cell phone for their own (and presumably ISAF’s) purposes. In taking these steps, TF81 personnel were, we consider, acting under their powers under the relevant Security Council resolutions and International Humanitarian Law rather than in a law enforcement context. As we have explained, New Zealand was also responsible for transporting Miraj in their vehicles to the facility where he faced a real risk of torture. Shortly after the operation, a deployed NZSIS officer, who had been told that Miraj had made a confession, enquired of Afghan authorities when New Zealand might expect to receive a debriefing in relation to it. A copy of Miraj’s confession was provided a few days later. This suggests that Afghan authorities understood that New Zealand had a particular interest in Miraj and whatever information he divulged.
- We accept that Afghan personnel were involved in the operation, but only to the extent of entering the mosque, executing the arrest warrant and undertaking site exploitation. TF81 personnel were not with them at the time, but that was the result of the fact that it turned out that Qari Miraj was not in the compound where he was originally thought to be but in the mosque opposite. TF81 personnel did make a forcible covert entry into the compound when there was no response to the “soft knock” but ISAF operating procedures precluded that in the case of the mosque. The Crown Agencies argue that this is a decisive consideration: whatever might have been the position if Miraj had been in the compound opposite, the fact was that Miraj was in the mosque with his companions; TF81 personnel did not enter and so were not present when Miraj was arrested and site exploitation was carried out; consequently, Miraj was under the jurisdiction of Afghanistan, not New Zealand.
- We do not see it in that way, given the matters set out above. Looking at the operation overall, while Miraj was undoubtedly of interest to Afghanistan from a law enforcement perspective, he was of special interest to New Zealand as a result both of his leading role in the 3 August 2010 attack and the threat he posed to the NZPRT in the future. In reality, Operation Yamaha was:
- substantially aimed at protecting the NZPRT in Bamyan, albeit that it would also protect Afghan forces and the local population;
- initiated by New Zealand personnel through the intelligence obtained;
- planned and triggered by New Zealand personnel, with limited input from Afghan authorities;
- directed by New Zealand personnel;
- substantially conducted by New Zealand personnel, the exceptions being the “soft knock”, the execution of the search warrant and the site exploitation of the mosque.
- Further, New Zealand obtained the “fruits” of the operation (directly by search and seizure, or, in the case of the confession, indirectly) and exploited them. TF81 personnel had physical control over Miraj for an hour or more from the time TF81 took over from the Afghan officers who entered the mosque, to the time TF81 personnel left the NDS facilities to return to Camp Warehouse. During this period they not only had physical custody of him but took possession of personal information obtained from him (the biometric data), information stripped from his cell phone, information obtained as a result of searches (specifically some documents) and, later, information from his confession. It was TF81 personnel who transported Miraj to NDS 90 in a TF81 vehicle and physically handed him over to those running the facility. From a substantive perspective, we consider that New Zealand personnel, rather than Afghan officials, were ultimately responsible for delivering up Miraj to NDS custody at the facility.
- Accordingly, we do not agree with the argument of the Crown Agencies that New Zealand did not exercise “control and authority” over Miraj, so that New Zealand’s non-refoulement and related preventive obligations were not engaged. We consider they were engaged. As planned, the only substantive difference between Operation Yamaha and an operation carried out by TF81 personnel acting alone was that there were Afghan personnel present during Operation Yamaha who would conduct the “soft knock” and execute the arrest warrant against Miraj. That was their sole function; yet on the Crown Agencies’ argument that would be sufficient to affect fundamentally the obligations in relation to torture that would otherwise have applied to New Zealand. But such a result would not, in our view, reflect the reality of the operation, nor would it reflect the true nature of the relevant obligations. The fact that, as the operation played out, Afghan personnel also had to conduct the site exploitation does not cause us to change our analysis.
- Apart from this, we consider that there is a strong argument that New Zealand may at least be regarded as jointly responsible for the operation and the resulting placement of Miraj in custody—either for the duration of the operation, or at the very least for the period during which Miraj was under New Zealand’s control. NZDF personnel were entitled to exercise control and authority over Miraj by virtue of their mandate to operate in Afghanistan and their arrangements with Afghanistan. As we see it, Miraj was subject to New Zealand’s authority in more than a de minimis way for more than a de minimis period. In essence, New Zealand and Afghanistan divided the essential tasks of his capture and transfer between them and cooperated throughout to ensure that detention was the final outcome.
- New Zealand provided the impetus for the operation. The Afghan personnel involved could not have executed the warrants without substantial assistance from New Zealand. Where two states collaborate and share the task of returning and taking an individual into custody, it is possible for each state to bear separate (and, at times, overlapping) obligations towards that person. The Crown Agencies accepted this, but argued that it was not possible for two states to have primary responsibility for the breach of the same obligation in relation to the same person where jurisdiction is the basis for responsibility. It was argued that International Human Rights Law and International Humanitarian Law envisage only one detaining authority, and in this case that was Afghanistan.
- We are unable to accept the Crown’s argument. It is possible, both as a matter of logic and of legal principle, for two or more states to bear separate yet concurrent primary obligations towards an individual, with each state responsible to the extent that acts or omissions are attributable to it, and in light of the way in which those acts or omissions cause or contribute to the outcome at issue (for example, to the transfer of an individual, or to his or her subsequent treatment). In complex detention operations, each state enjoys authority to act (and thus to exercise jurisdiction) as a result of one or more sources of legal authority.
- With legal authority comes responsibility: states bear obligations which derive from various sources of law and attach to the state when it operates abroad. These obligations may derive from treaty provisions applied extraterritorially (such as the Convention against Torture); specific treaty obligations that, by their nature, may apply without territorial limits (such as Article 1 and Article 3 of the Geneva Conventions on ensuring respect and the protection of detainees); customary international law obligations; and obligations flowing from peremptory norms (such as the obligation to prevent torture through a range of effective measures).
- In short, obligations may take different forms when applied abroad, but they follow a state when it engages on the international stage and apply alongside the obligations of others. States are not relieved of their individual responsibilities by virtue of their role in discharging international mandates (including Security Council mandates). Similar reasoning may be applied when the implementation of such mandates involves leading or assisting in joint operations with other states.
- In partnered operations and similar contexts, jurisdiction may not, and need not necessarily, be exclusive. Nor is it necessarily the case that a form of jurisdiction held by one state ‘extinguishes’ the jurisdiction of others. This is particularly relevant in contexts such as counter-terrorism and trans-national crime, where transfers are often the result of coordinated, incremental steps by multiple states to effect a transfer (whether lawfully or otherwise). When two or more states exercise different forms of jurisdiction over people, places or situations, therefore, it is useful to consider whether one of these forms of jurisdiction was the most relevant—and, if so, how or why it was relevant. When examining jurisdiction, context is important. Care should be taken to consider whether a state’s conduct (including its handling of a detainee) had a decisive, or at least a material, impact on the particular right or obligation at issue or on the outcome or issue being examined (here, the transfer of a person from a place of safety to a place of risk).
- In Operation Yamaha, Afghanistan authorities lawfully exercised an arrest warrant and Miraj was clearly subject to it. It provided a basis for his subsequent entry into custody, interrogation, and possible trial (although, as noted elsewhere, the arrest warrant was not essential: NZDF had lawful authority to detain him on its own had it elected to do so).
- Despite Afghanistan’s de jure control via the initial arrest, from the moment New Zealand uplifted Miraj from Afghan authorities he was under their effective control and authority. While he was in their physical custody, they also exercised legal powers derived from the Security Council mandate to take and keep Miraj’s personal data (biometric information, the contents of his cell phone and documents) before they transported him in their vehicles and handed him over to NDS at NDS 90. New Zealand’s handling of Miraj was more than an incidental form of assistance: it was the final act and the active part of the transfer, and it followed a series of other essential contributions by TF81 over a longer period to remove Miraj from the battlefield, to process him and to secure his detention. New Zealand had an interest in that detention in light of the intelligence that might be gleaned from the ensuing investigation.
- We consider, therefore, that the operation and transfer engaged New Zealand’s non-refoulement and related preventive obligations arising under the Convention Against Torture, customary international law, and as a consequence of the peremptory norm prohibiting torture. This is because, in truth, the operation was New Zealand-led and because TF81 exercised control and authority over Miraj during the relevant part of the operation.
- In addition, we accept that New Zealand was obliged under Common Article 1 to the Geneva Conventions to exercise due diligence to ensure that Afghanistan complied with its obligations under Common Article 3 in relation to the treatment of detainees, which is supported as well by New Zealand’s other preventive obligations in relation to torture. In fulfilling these obligations, New Zealand could have taken steps at a systemic level and at an individual level (ie, in relation to Miraj specifically). As we will discuss below, we believe that New Zealand was obliged to take steps specifically in relation to Miraj.
- To summarise, Qari Miraj was an insurgent in whom New Zealand had a particular interest; New Zealand was the prime mover behind the operation to capture him; New Zealand played the leading role in organising, triggering and executing the operation, and would have played an even greater role had Miraj not been in a mosque; and New Zealand took responsibility for handling and processing Miraj before he was handed over to the NDS at NDS 90, obtained personal and other information from him, and later exploited that information. New Zealand clearly had an interest in whatever information Miraj divulged in detention. Given those circumstances, we think it clear that New Zealand had an obligation to satisfy itself as to the conditions of Miraj’s detention. The question now is whether it took sufficient steps to do so.
Did New Zealand breach its obligations?
- The Crown Agencies argued that, if New Zealand’s non-refoulement obligations were engaged in Operation Yamaha, New Zealand was not aware of the risks that Miraj faced in Afghan detention given the information available at the time. Accordingly, New Zealand was not in breach of its obligations.
- The Crown Agencies emphasised that:
- ISAF considered that that transfers to NDS facilities in Kabul were permissible, as did other ISAF partners.
- The knowledge required is knowledge relating to the risks faced by a particular individual at a particular facility. New Zealand had no knowledge of any risks Miraj would face at NDS 90, especially given that the facility which caused concern in Evans was NDS 17.
- New Zealand authorities did not appreciate at the time of Operation Yamaha that NDS 90 was in the same compound as NDS 17 and that detainees could be transferred from one facility to the other.
- New Zealand was not aware of any evidence of widespread torture by the NDS in detention facilities over the period 2010–2011 until shortly before the UNAMA report was released in October 2011. The Crown Agencies acknowledged, however, that New Zealand could have undertaken further investigation as to conditions in Afghan detention facilities, as the Solicitor-General had recommended.
- We see these points as being relevant to both the non-refoulement and Common Article 1 obligations just discussed.
- We have seen no evidence that New Zealand made any enquiries specific to Miraj. Rather, New Zealand relied on:
- the view of ISAF and other contributing nations that it was appropriate to send people to detention in NDS facilities in Kabul;
- what the Minister was told in discussions with Afghan counterparts in August 2010, to the effect that detention conditions were improving and there was greater scrutiny of NDS detention facilities;
- meetings with NDS officials after the Evans decision to obtain assurances that the issues raised in Evans were being addressed; and
- interactions with the ICRC about detention issues in Afghanistan.
- As can be seen, New Zealand did not itself undertake any rigorous enquiry into conditions in Afghan detention facilities, or even attempt to reach an independent view about them. Rather, New Zealand relied on the views of others and assurances given in discussions with Afghan authorities, including the NDS itself. It did this despite the Solicitor-General’s opinion that New Zealand had a duty to, among other things, gather information about the practices of Afghan personnel and institutions. As we understand it, the Crown Agencies accept that New Zealand officials should have undertaken a more rigorous and independent information-gathering process to satisfy themselves that the information received was accurate.
- We think it disappointing that New Zealand authorities did not take greater steps at the time of the deployment in 2009 to find out what current conditions in Afghan detention facilities were, given that they were aware that international reports had identified significant problems over a number of years. It was improbable that conditions in Afghan detention facilities would have improved dramatically over a relatively short time, even those in Kabul. The effect of the Evans decision was to confirm that the problem of detention conditions persisted. The judgment in that case cited various United Nations documents and reports by reputable NGOs and other bodies. One such report, from Human Rights Watch in December 2009 and its country summary for 2010, stated that there were persistent reports of torture and abuse of detainees held by the NDS. Similar reports cited in the judgment included those from the United Nations High Commissioner for Human Rights in January 2009, Afghanistan’s report to the Human Rights Council as part of the Universal Periodic Review in February 2009, the US State Department in March 2008, and the 2008 Annual Report of the UK Foreign Affairs Committee of the British House of Commons. The Solicitor-General’s opinion was unequivocal in stating that New Zealand had an obligation to gather information.
- In these circumstances, we were unimpressed with the Crown Agencies’ submission that there was no evidence known to New Zealand of widespread torture by the NDS in detention facilities over the period 2010–2011 until shortly before the UNAMA report was released in October 2011. The reality is that New Zealand did not look very hard, despite what we consider to be obvious indications that torture and mistreatment remained prevalent. Indeed, in light of the obligation to investigate identified by the Solicitor-General, which New Zealand did not fulfil adequately, the argument has something of the quality of “seeking to take advantage of one’s own wrong”.
- Further, we have the impression that TF81 personnel took little interest in which facilities the CRU or other Afghan forces would send those they arrested on partnered operations to, on the basis that it was none of New Zealand’s business—not surprisingly perhaps, given the Government’s policy on detention during partnered operations.
- New Zealand would be in breach of its non-refoulement obligations if there were substantial grounds to believe that Miraj faced a real risk of torture if it handed him over to the NDS at NDS 90. New Zealand made no particular enquiries about that as it did not consider that it was under any obligation to do so—at most, it was a discretionary matter. While we accept that ISAF favoured the use of NDS detention facilities in Kabul for operational reasons, we do not think it acceptable that New Zealand authorities did not realise that NDS 90 was in the same compound as NDS 17 and that detainees could be moved between the two facilities. Evans should have resulted in meaningful enquiries. In the result, we consider that New Zealand is not entitled to rely on its lack of information to argue that it did not breach its non-refoulement obligations. On an objective view, we consider it is clear that, as at January 2011, there were substantial grounds to believe that detainees such as Miraj faced a real risk of torture or mistreatment at NDS 90.
- The same analysis applies in respect of New Zealand’s obligations under Common Article 1 and Common Article 3. We consider that New Zealand did not take sufficient steps at a systemic level to meet its obligations under those Articles. In addition, given New Zealand’s particular interest in Miraj and the other aspects of Operation Yamaha discussed above, we consider that New Zealand had an individual obligation to Miraj and was obliged under the Common Articles to satisfy itself that he did not face a real risk of mistreatment in detention. This required active steps on New Zealand’s part, which it did not take.
- The Crown Agencies argued that the Inquiry is, unfairly, applying “hindsight analysis” and assuming that if New Zealand had made further inquiries, it would have discovered what UNAMA found. We do not accept this:
- First, New Zealand was sufficiently concerned about conditions in Afghan detention facilities that it considered it necessary in 2009 to enter into a written arrangement with Afghanistan relating to the treatment of people NZDF personnel detained on non-partnered operations and transferred to Afghan control. It was well aware that there was a problem that needed to be addressed.
- Second, Evans provided clear warnings in the mid-2010 that conditions in NDS detention facilities had not improved dramatically so as to remove the risk of torture or mistreatment in detention.
- Third, a complaint of this nature does not sit well, given that Crown Agencies accepted that New Zealand did not meet the investigative obligations that the Solicitor-General had identified.
- Given the views we have expressed, we do not need to address the issue of complicity here, although we dealt with it briefly in our outline of the law in chapter 10, given its importance. However, we should record that we accept that New Zealand authorities did not intend for Miraj to be tortured, nor have we seen any evidence that they had actual knowledge of any intent on the part of Afghan authorities to torture him. We leave open the position in relation to constructive knowledge.
New Zealand’s response to the allegations of torture
- As we have described, New Zealand authorities received a copy of Qari Miraj’s confession not long after it was made. About a month later, they learnt of the allegations that he had been tortured. Despite the policy position which we quoted at paragraph , no one appears to have taken any steps with Afghan authorities or counterparts to establish what, in fact, happened to Miraj; to express New Zealand’s clear position on torture and concern about the allegations through appropriate channels; to brief senior leaders (internally) or relevant ministers (for whom this information would have been highly relevant given the decisions they were being advised to make); or to review New Zealand’s legal and policy position as it then stood. In interviews with the Inquiry, it was clear that officials appreciated the seriousness of torture as a legal or moral issue; however, none were able to explain in detail the policies or procedures that existed for escalating, investigating or following up allegations of torture.
- In his advice, the Solicitor-General recalled the Committee against Torture’s guidance which states parties to the Torture Convention that become aware of torture in the course of joint operations “are then obligated to report and seek investigation of those allegations.” He noted that such information might change the overall assessment of New Zealand’s legal responsibility for persons detained on Afghan-partnered operations.
- Given that advice, it was essential that New Zealand have clear and effective policies in place so that this form of evaluation, both proactive and reactive, could take place. Against the backdrop of international and United Nations reporting that existed up to and including that time, the information on Miraj’s treatment was specific and highly relevant to New Zealand’s position. It was, in short, an allegation made in the context of an operation where New Zealand had played a leading role in successfully delivering up a high-value target to Afghan intelligence officials. Ministers ought to have been briefed and provided with an opportunity to reconsider the legal and policy position and direct officials on next steps. Afghanistan was under an obligation to investigate any allegation of mistreatment, even if not made directly to its authorities. New Zealand, as the state that, in substance, led the operation and played a material role in Miraj’s transfer to NDS, was under a similar obligation to follow up, and to consider and, if necessary, revise its overall approach.
- The Crown Agencies accepted in submissions that New Zealand did not meet its obligations in this respect and acknowledged that further investigation may be needed or additional training could be provided to operational officials so that they recognise where such allegations need to be reported, and what further steps should be taken. We agree.
- We should also note that the IGIS has addressed this issue in some detail in her report. There are two features of her analysis to which we should draw attention.
- First, the Inspector-General records that on 3 February 2011, relevant ministers were informed that Miraj and others had been arrested and that Miraj and one other person remained in detention in Kabul. A deployed NZSIS officer in Kabul advised NZDF, the NZSIS in Wellington and the New Zealand Ambassador in Kabul that news of Miraj’s detention had spread throughout Tala wa Barfak and that his fellow insurgents were focused on obtaining his early release by whatever means necessary, including through the Afghan “reconciliation” process. Given New Zealand’s special interest in Miraj, the deployed NZSIS officer sought and received assurances from the NDS that Miraj would not be released but would remain in detention. The Inspector-General describes this as consistent with “the New Zealand Government stance”. The concern that Miraj remain in custody continued even after New Zealand agencies learnt of the allegations that he had been tortured.
- Second, the Inspector-General examines in some detail how the intelligence agencies responded to the allegations that Miraj had been tortured. She identifies faults with the way the matter was handled, particularly because New Zealand had urged the NDS to keep Miraj in custody. She sees this as giving rise to a duty on the New Zealand Government, through its officials, to assure itself that there was no real risk that Miraj was being tortured or mistreated. She expressed concern that the NZSIS did not identify any responsibility to make sure the Government understood the conditions of Miraj’s detention and to assess whether they were appropriate in light of New Zealand’s legal obligations. As will be obvious, we share the IGIS’s concern.
Conclusion and findings
- When the New Zealand Government decided to deploy the NZSAS to Afghanistan in 2009, it followed ISAF’s policy that Afghan authorities would be responsible for persons arrested and/or detained by Afghan personnel on partnered operations; ISAF personnel would only become responsible for such people if they assumed control and placed them in detention. NZDF personnel followed that policy and also developed the arrest warrant mechanism. This mechanism was, as we understand it, ultimately adopted more widely within ISAF. Moreover, the Government’s detention policy was supported in principle by an opinion from the Solicitor-General.
- In our view, the difficulty with the policy, as it was applied, is that it did not take sufficient account of the facts of particular operations—in particular, the comparative roles of NZSAS and Afghan personnel in those operations. Rather, the policy gave decisive weight to form. In general, it treated as decisive the fact that an Afghan official executed an Afghan arrest warrant to arrest the target. The policy did so even though these warrants were generally prepared by NZSAS personnel—in some cases, after substantial involvement by NZSAS personnel in making the case for arrest (for example, by developing admissible evidence from intelligence) and even though the operation was substantially instigated, planned, organised and carried out by New Zealand personnel. In respect of operations such as Operation Yamaha, this approach seems to us to have the appearance of a “workaround”.
- The policy as applied meant that there was little real incentive for New Zealand authorities to find out exactly what the conditions were in NDS detention facilities, given that they saw New Zealand’s non-refoulement and similar obligations as arising only in relation to very few detainees, in respect of whom there was a written arrangement with Afghanistan, which included a monitoring regime. This was despite the fact that the Solicitor-General had advised that New Zealand had a duty to undertake enquiries as to the conditions of detention in the context of his discussion of partnered operations.
- Given publicly available information about conditions in Afghan detention facilities at the time of the deployment in 2009, including in relation to the NDS, we consider that New Zealand should have attempted to enter into some form of arrangement with Afghanistan, covering the treatment of persons captured during partnered operations. An individual monitoring regime in relation to each detained individual may not have been necessary (although it may have been necessary or appropriate in relation to some, such as Miraj, depending on the circumstances of the detention).
- What was required was some mechanism by which New Zealand could obtain accurate information about the treatment of people in Afghan detention facilities, with the ability to respond effectively (for example, by suspending cooperation) if conditions were found to be unsatisfactory. The Solicitor-General had, after all, said that New Zealand had a duty to obtain such information. The Government chose not to attempt to enter into an arrangement, however, or to take effective steps to obtain accurate information about conditions in Afghan detention facilities. Steps such as minister-to-minister conversations and discussions with Afghan authorities such as the NDS were simply not sufficient.
- We were particularly troubled by the weight that seems to have been given to resource implications in Government’s consideration of this issue, and by the attempts made in submissions to the Inquiry to justify New Zealand’s approach on the basis of lack of resources. Compliance with human rights standards requires a level of resource commitment that reflects the fundamental nature of values at stake and the harm that is to be prevented. We see in this an important issue going to New Zealand’s core values. Is it consistent with the core values to which New Zealand says it is committed to follow a policy that means it has no responsibility for taking steps to ensure appropriate conditions of detention for persons who its forces were closely involved in capturing, in part because it would require too much in the way of resources? If New Zealand is not in a position to commit the resources necessary, should it become involved in the enterprise in the first place?
- In relation to Operation Yamaha, we consider that, given the circumstances of the operation discussed above, Qari Miraj came within New Zealand’s “jurisdiction”, so that New Zealand owed certain obligations to him. The fact that he was arrested under an Afghan arrest warrant by Afghan officials does not mean that Miraj was not also within New Zealand’s jurisdiction for a period. New Zealand owed non-refoulement and related preventive obligations to Miraj; Afghanistan had obligations not to torture or mistreat him in detention.
- We consider that New Zealand was in breach of its obligations to Miraj because, at the time he was handed over to NDS custody, there were substantial grounds to believe that he faced a real risk of torture or other mistreatment given the information available about conditions in Afghan detention facilities, including NDS facilities in Kabul. New Zealand was in breach of the duty identified by the Solicitor-General because it had failed to take effective steps to find out about conditions in such facilities. There were also no enquiries in relation to Miraj’s position specifically.
- Further, we are satisfied that New Zealand breached its obligations under Common Articles 1 and 3 of the Geneva Conventions. New Zealand’s obligation was to “ensure respect” “in all circumstances” for the Conventions’ obligations, including in relation to the treatment of detained persons such as Miraj. In relation to the detention of Miraj, given the background to the operation, its nature, New Zealand’s role in it and New Zealand’s particular interest in information obtained from Miraj, New Zealand had a specific obligation to take all steps reasonably possible to ensure that Afghanistan complied with its obligations towards Miraj while in detention. New Zealand did not do so and was therefore in breach of its obligations under the Common Articles.
- Moreover, having learnt of the allegations of Miraj’s torture, New Zealand failed in its obligation to respond by making further investigations, including taking the matter up with Afghan authorities. Indeed, it does not appear agencies advised ministers of the allegation, which is obviously unacceptable given the circumstances of the operation and the nature of New Zealand’s interest in Miraj. As we understand it, the Crown Agencies accept that New Zealand did not respond to the allegations in the way it should have.
- We make two final points. First, as we noted during our brief overview of developments in relation to New Zealand’s detention policy, there were points at which officials could have brought particular matters to ministers’ attention (such as new information or developments, or the nature of TF81 operations with NDS) but did not. Obviously, it is a matter of judgement as to when something needs to be brought to a minister’s attention. But issues concerning New Zealand’s obligations in relation to torture in the context of its participation in a non-international armed conflict do engage the country’s interests in a fundamental way. Our impression, admittedly on the basis of our exposure to specific issues and evidence, is that ministers would have benefitted from learning of some of these matters when they arose.
- Second, one of the lessons that emerges from our analysis is that the way an issue is framed often determines how it is answered. A lawyer who is asked what New Zealand’s obligations are in relation to torture when NZDF personnel are “in the vicinity” when Afghan personnel arrest a suspect on a partnered operation is likely give a different answer than a lawyer asked to express a view about those obligations in relation to an operation such as Operation Burnham, which was, in substance, a New Zealand-led ISAF operation rather than a true partnered operation.
- The Crown Agencies repeatedly described partnered operations as operations in which persons were detained by Afghan authorities “with the support of ISAF forces”. A variant seen in contemporaneous documents was that the CRU took the lead on all operations and were accordingly the detaining authority. In respect of the operations we have examined closely—Operations Burnham, Nova and Yamaha—such descriptions do not accurately reflect what was planned or in fact occurred. The short point is that facts matter. This is particularly so when legal advice is sought. In that context, it is dangerous to talk in abstract generalities—rather, the legal adviser needs to be advised of the range of likely factual scenarios.
- It will be recalled that cl 7.9 of the Inquiry’s Terms of Reference requires us to examine:
Whether the NZDF’s transfer and/or transportation of suspected insurgent Qari Miraj to the Afghanistan National Directorate of Security in Kabul was proper, given (amongst other matters) the June 2010 decision in R (oao Maya Evans) v Secretary of State for Defence  EWHC 1445;
- Our answer is that we consider that in handing Miraj over to the NDS at NDS 90, New Zealand breached its non-refoulement and related preventive obligations, and its obligations under Common Article 1 of the Conventions to “ensure respect” for the obligations contained in Common Article 3 concerning the treatment of detainees. The hand-over was accordingly improper.
- Arising out of this, we have some recommendations to make on the matter of detention, which we will discuss in chapter 12.
Timeline of key events relating to detention and Operation Yamaha
Credible reports of torture in Afghan government law enforcement facilities are published by a number of international bodies, rapporteurs and NGOs.
The New Zealand Government enters an Arrangement with the Government of Afghanistan providing that detainees captured by TF81 and transferred to Afghan custody will be subject to monitoring and other checks to ensure their human rights are respected.
NZDF considers whether reporting to ISAF on detentions carried out by Afghan partners and collecting biometric data of detainees would mean that New Zealand incurred responsibilities for detainees. It concludes these steps would not mean suspects were “detained” by TF81 personnel. Officials also provide other advice that considers whether to pursue assurances in relation to detainees captured during partnered operations.
The Evans case identifies specific Afghan National Directorate of Security (NDS) facilities where detainees are at risk of torture.
NZDF Legal Services and the Solicitor-General consider the implications of Evans on New Zealand’s detention policy in Afghanistan. The Minister of Defence visits Kabul in August and discusses detention with Afghan officials and others.
TF81 receives information that insurgent leader Qari Miraj, linked to the 3 August 2010 attack on the NZPRT, will be transiting through Kabul. An operation working with personnel from the NDS to capture him is urgently planned and approved. TF81’s usual partner unit, the Crisis Response Unit, does not participate in the planning or conduct of the operation.
Qari Miraj and four associates are located in a mosque in Kabul in the early hours of the morning. As ISAF personnel are not permitted to enter mosques, NDS personnel and local police enter and find Miraj and his associates. The men surrender peacefully and are taken by TF81 to the NDS facility in Kabul.
An NZSIS employee stationed with TF81 in Kabul receives a copy of Qari Miraj’s confession from the NDS.
The NZSIS disseminates the confession to a number of other New Zealand government agencies.
New Zealand officials in a number of agencies learn of allegations that Miraj was tortured to extract a confession.
A senior official in one agency outside NZDF files the allegations of torture in a “general oversight file”. No other action is taken.
ISAF amends its detention policy in anticipation of a United Nations Assistance Mission in Afghanistan (UNAMA) report on torture in Afghan government detention facilities, adopting a number of measures to address the issue.
UNAMA releases its report Treatment of Conflict-Related Detainees in Afghan Custody, detailing widespread allegations of torture and inhumane treatment in Afghan detention facilities, including NDS facilities.
Officials brief ministers on the UNAMA report and the actions ISAF is taking. The Minister of Defence is advised that New Zealand’s original detention policy has been reviewed, is robust and can deal with the issues raised by UNAMA.
New Zealand is informed of further allegations of torture made by detainees in NDS facilities.
< Return to Contents | Previous | Next >