Challenging Times Ahead: Australias War Crimes Prosecutions

On Tuesday, 7 April 2026, Australia's most decorated living soldier, Ben Roberts-Smith, was arrested and charged with five counts of the war crime of murder. He is the second soldier charged following Australia's Brereton Inquiry, which, in November 2020, found credible information of 23 incidents of unlawful killing and two incidents of cruel treatment by Australian Special Forces soldiers in Afghanistan between 2009 and 2013 (for further detail see here). The other person against whom charges have been laid, Oliver Schulz, was arrested in March 2023 and committed to trial in August 2025. Schulz has pleaded not guilty to the crimes. The trial is currently expected to commence in 2027. 

Domestically, these trials raise complex legal, evidentiary and procedural issues that will test Australia's capacity to prosecute international crimes. More broadly, this represents a key opportunity to explore the distinct challenges raised by prosecuting international crimes in a common law system.

Trial Modalities and the Right to a Fair Trial

Section 80 of the Australian Constitution requires that ‘[t]he trial on indictment of any offence against any law of the Commonwealth shall be by jury'. The High Court of Australia has interpreted section 80 to provide an absolute right to a trial by jury that cannot be waived by a defendant. This means that all trials of alleged war crimes must take place before a jury, a distinctive element of a common law system. Under the Australian system, for prosecution of Commonwealth crimes, a jury decision to convict must be unanimous.

The involvement of the jury raises several questions. Allegations surrounding war crimes in Afghanistan have been the subject of significant media attention since 2017 when the ABC published the Afghan Files, reporting on a series of leaked documents including video footage of possible unlawful killings. More recently, the allegations against Roberts-Smith were the subject of significant media attention during a defamation case, unsuccessfully brought by Roberts-Smith against media outlets who reported allegations relating to his alleged involvement in the murder of unarmed civilians while serving in Afghanistan. After a lengthy trial, Justice Besanko, in the Federal Court of Australia, found that the media outlets had proven – on the balance of probabilities – the truth of their claims that Roberts-Smith had committed war crimes in Afghanistan. The decision was confirmed on appeal by the Full Federal Court in May 2025, with the High Court of Australia refusing special leave to appeal in August 2025 (see A proxy war crimes trial).

The amount of media attention that the issue of Australian war crimes in Afghanistan has attracted – and the fact that Roberts-Smith has already lost a civil case, albeit one where the standard of proof is the balance of probabilities – raises questions about the extent to which jurors can and will approach criminal proceedings with an objective and impartial mind. In the most extreme cases, prejudicial pre-trial publicity might result in an accused applying for a stay of proceedings on the basis that the accused can no longer receive a fair trial. 

With criminal proceedings now on foot, care will need to be taken to avoid material entering the public domain via media reporting or other means that might taint the jury pool. One way that this could be managed is by using suppression and non-publication orders. On the other hand, Australia respects the principle of open justice, and there will be significant public interest in following the outcome of this case. The judges assigned the case will have to navigate the balance between ensuring the integrity and fairness of the proceedings and the extent to which access can be open. 

Domesticating International Criminal Law

The law on which the charges are based has never been used, and there are questions as to its interpretation. Both Schulz and Roberts-Smith are charged with murder as a war crime under section 268.70 of the Commonwealth Criminal Code. This provision represents Australia's domestic implementation of Article 8 of the Rome Statute; however, it is not a direct translation and varies in some significant respects. The equivalent provision in the Rome Statute is Article 8(2)(c)(i), which criminalises ‘any of the following acts committed against persons taking no active part in the hostilities, including members of the armed forces who have laid down their arms and those placed hors de combat by sickness, wounds, detention or any other cause: (i) Violence to life and person, in particular murder of all kinds…'. While aspects of Article 8(2)(c)(i) and the corresponding elements of the crime are reflected in section 268.70, there are key textual differences between the international definition of the crime and that found in the Criminal Code. In particular, section 268.70 seeks to exclude ‘incidental' or ‘collateral' deaths of civilians from its scope; namely, civilians injured incidentally to an attack on a military objective. This rather confusingly overlaps the provision with another distinct war crime found only in international armed conflicts. This overlap is not made at the international level. Another key difference is the requirement that the victim is not taking an active part in hostilities nor is a member of an organised armed group, which excludes certain people from the category of war crime victims by virtue of their membership of a group (for example, the Taliban) rather than their actions. The provision then removes from the scope of the exclusion members of a group who are hors de combat. The provision is — at best — clumsily drafted and overly complex, with layers of exclusions, particularly when compared with the relative simplicity of the Rome Statute equivalent. 

Another key deviation from the Rome Statute is the decision not to incorporate the ‘general' provisions of the Rome Statute into Australian law, such as Article 30 on the mental elements required for ICC crimes. Instead, recklessness has been included in section 268.70 alongside knowledge as a means of establishing the mental element; the Rome Statute does not include recklessness. Further, as the offence is situated within the broader context of the Criminal Code, the general definitions in Division 3 of the Code will apply; this includes provisions on fault elements, particularly section 5.3 (knowledge) and 5.4 (recklessness).  With the exception of command responsibility, the ICC modes of liability have not been directly translated into Australian law, meaning that prosecution must rely on the modes of liability found in the general provisions of the Criminal Code. For Schulz, this only includes direct perpetration. For Roberts-Smith this is direct perpetration – intentionally causing the death of a person – for two charges, and a form of complicity or common purpose liability under section 11.2 of the Criminal Code – aided, abetted, counselled or procured another person to intentionally cause the death of a person – in respect of three charges, which relate to allegations of ‘blooding' new soldiers. Command responsibility is not raised on the facts.

Evidentiary Issues

The presence of a jury will impact the approach to evidence adopted by the court. The position in NSW resembles the position in international courts, with evidence generally admissible unless an exclusionary rule applies. However, given the presence of the jury, the parties — particularly the defence — are conscious of the risk of evidence prejudicing the jury and can be expected to apply evidentiary rules rigorously.

One immediate question is whether the extensive evidence obtained by the Brereton Inquiry will be admissible at trial. The short answer is that material from the Inquiry will not be admissible due to the use and derivative use immunity extended to those who gave evidence to the Inquiry. This immunity means that any document produced or information given or produced by a witness or obtained as a direct result of the witness's participation are not admissible against the individual in any civil or criminal proceeding in any federal court or State court or proceedings before a service tribunal. The only exception is in relation to proceedings concerning the giving of false testimony. The grant of immunity was considered appropriate as the Brereton Inquiry had powers of compulsion and could compel witnesses to appear. To manage the flow of information between the Brereton Inquiry and prosecution authorities and prevent a prosecution becoming tainted through reliance on immune evidence, the Australian Government established the Office of the Special Investigator (OSI), in November 2020. The OSI is an independent agency within the Commonwealth Attorney-General's department that works with the Australian Federal Police (AFP) to gather evidence of alleged war crimes committed by Australian forces in Afghanistan. However, the challenges in managing the flow of information from the Brereton Inquiry and the complex nature of the applicable immunities are likely to cause significant delays in the criminal justice process, and is one of the reasons why Schulz, who was arrested and charged in March 2023, is yet to be put on trial. Information and potential evidence must be released by the Inquiry to the OSI, who reviews for any immune evidence before filtering that information to the AFP, the Commonwealth Director of Public Prosecutions and the defence lawyers.

These cases are different to what we see in universal jurisdiction cases in other countries. This is Australia prosecuting its own troops for war crimes. Much of the evidence will be provided by the Australian Defence Force (ADF) itself and/or raise potential classified defence documents or information. As a result, both proceedings are subject to the application of the National Security Information (Criminal and Civil Proceedings) Act 2004 (Cth) (‘NSICCP Act'). The object of the Act is ‘to prevent the disclosure of information in federal criminal proceedings and civil proceedings where the disclosure is likely to prejudice national security, except to the extent that preventing the disclosure would seriously interfere with the administration of justice'. Orders will need to be made – usually by agreement between the parties and the Commonwealth – to manage the proceedings and to regulate access to the sensitive evidence. Agreeing those orders takes considerable time, creating extensive delay and directly impacts the length of proceedings. There has been a strong Commonwealth presence throughout the committal proceedings in Schulz, with the Commonwealth effectively policing the consent orders, intervening whenever topics touch on national security information. Sensitive topics included information as to the ADF rules of engagement, specific tactical engagement practices, details of equipment, oversight and partner force capacity, and standard ADF operational practices. Two courts were made available: a public court that could be closed on occasion and a fully closed court, with access limited only to the accused and lawyers with pre-approved security clearance. Proceedings shifted between open and closed court as required. 

The NSICCP Act has been widely criticised as being too focused on protecting national security information at the expense of open justice and fair trial rights. In the Schulz proceedings, Justices Hamill and Yehia in the Supreme Court have recognised the tension between giving effect to the NSICCP Act and ensuring an expeditious trial. What we have seen so far in Schulz suggests that significant parts of both trials will be held in closed court, for which no transcript or media reporting will be available. This will limit the capacity of the trials to provide a complete account of what happened in Afghanistan, both to a domestic and an international audience. Given the limited public information available about other ‘secret' trials where the Act has impacted public access, it is also hard to imagine how the Act will apply in a trial by jury. Will national security information be provided to — or kept from — the jury? Will jury members require security clearance? 

Even with the operation of the NSICCP Act, there is the potential for a public interest immunity application by the Commonwealth. A successful claim would result in the Commonwealth being able to withhold immune material from not only the public, but also the court, the parties and the jury. This means neither party would be able to see or rely on that information, which has obvious implications for defendants' right to a fair trial, as the withheld material may be inculpatory or exculpatory. One of the aims behind the introduction of the NSICCP Act was to limit the scope for such broad immunity claims; however, as a common law action, public interest immunity remains as an option alongside management of information through the Act. Several claims for public interest immunity were made during the Roberts-Smith defamation trial. The Commonwealth in Schulz signalled a potential application for public interest immunity during the committal hearing but it is not clear if this claim was made and determined, as everything concerning this application was done in closed court. Both prosecution and defence could be expected to object to any such claim, noting the detrimental impact for the trial — and public confidence — that a successful claim may have. 

Conclusion

This may only be the start of Australia's engagement with international criminal law domestically: the OSI have indicated that a further 10 investigations are still ongoing. The proceedings in Schulz so far illustrate clearly the challenges of translating the Rome Statute into Australian law, in particular the differences between the domestic legal framework and its common law foundations and international standards and practices. The legal issues and lengthy delays show that domestic prosecutions of international crimes are usually not easy, even in well-resourced States, prosecuting their own nationals and where the State controls much of the evidence. Challenges with domestic prosecutions makes it harder for Australia – and other states – to criticise the ICC but also means we must be clear as to how we measure success for domestic prosecutions. What constitutes success is not only the outcome; Australia has not ‘failed' if a trial results in an acquittal. Rather, Australia must demonstrate a genuine commitment to criminal prosecution, a good faith performance of its duty to prosecute and investigate international crimes committed by its forces. With many European countries successfully conducting universal jurisdiction prosecutions, Australia's cases represent an opportunity to learn from these experiences, but also to explore the unique legal and procedural challenges of international crimes proceedings in a common law system.