Federal Court of Australia

R&B Investments Pty Ltd (Trustee) v Blue Sky Alternative Investments Limited (in liq) (Security for Costs and AI) [2026] FCA 1265

ORDERS

THE COURT ORDERS THAT:

1.    On or by 8 October 2026, the Applicants are to provide security in the sum of $680,000 to BSAIL for its costs up to the conclusion of the conferral the subject of order 4 made on 17 July 2026.

2.    On or by 8 October 2026, the Applicants are to provide the following security to EY:

(a)    $600,000 in respect of EY’s past costs incurred to 7 August 2026; and

(b)    $680,000, in respect of EY’s costs incurred from 8 August 2026 up to and including the completion of the conferral the subject of order 4 made on 17 July 2026.

3.    The security referred to in Orders 1 to 2 above be provided by way of:

(a)    payment of cash into Court;

(b)    bank guarantee;

(c)    a deed of indemnity issued by an insurer agreed by the relevant respondent (or determined by the Court), in a form reasonably acceptable to the parties (or determined by the Court); or

(d)    such other form of security as may be agreed between the parties (or determined by the Court).

4.    Pursuant to r 2.42(2)(a) of the Federal Court Rules 2011 (Cth) (FCR), any cash security paid into Court be transferred to an interest-bearing bank account established and operated by the Court.

5.    If security is not provided in accordance with Orders 1 to 2 above (or any part thereof), the proceeding be stayed pursuant to r 19.01(1)(c) of the FCR as against the First to Fourth Respondents.

Conferral

6.    By 6 November 2026, pursuant to s 37P(2) of the Federal Court of Australia Act 1976 (Cth), the parties are to:

(a)    confer, with the assistance of a suitably qualified technology experts selected by the parties, as to the suitability of using artificial intelligence (AI) technology in the preparation of the matter for mediation and trial including, without limitation, the parties conferring regarding the use of AI technology in:

(i)    identifying, collating and synthesising information relevant to the facts in issue and identifying areas of agreement and disagreement between the parties for the purposes of simplifying the dispute;

(ii)    preparing chronologies and summaries of primary documents, including by reference to specified issues; and

(iii)    managing and interpreting documents and other data that is relevant in the proceeding more generally;

(b)    produce a report and/or draft protocol to the Court as to the results of that conferral;

(c)    have reference, to the extent necessary, to the Federal Court of Australia, General Practice note - Use of Generative Artificial Intelligence Practice Note (GPN-AI), 16 April 2026, during that conferral.

Note:    Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

REASONS FOR JUDGMENT

(Delivered ex tempore, revised from the transcript)

LEE J:

A    INTRODUCTION

1    The applications before me are applications for security for costs pursuant to s 56 of the Federal Court of Australia Act 1976 (Cth) and r 19.01 of the Federal Court Rules 2011 (Cth).

2    This is a highly complex class action. Indeed, the applicants have filed something in the region of 1,400 pages of documents seeking to articulate the case they propose to prove at trial. Because of a range of interlocutory disputation, the proceeding, which commenced in 2022, has not moved at the pace of summer lightning.

3    We have now reached the stage where, by the end of the year, there is to be a structured conferral process directed to identifying what facts, out of the universe of potentially relevant facts, can either be agreed or at least treated as non-contentious (in the sense that a party will not seek to dispute the fact at trial). For reasons unnecessary to detail for present purposes, that process assumes particular importance in a case such as this, which will be extraordinarily document-heavy.

4    Even though no lay or expert evidence has yet been filed by the respondents, it appears that the following amounts have already been expended by them on legal costs: Blue Sky Alternative Investments Limited (Blue Sky), $1.2 million; the second respondent, Mr Robert Shand, $2.5 million; the third respondent, Mr John Kain, $1.6 million; and the fourth respondent, Ernst & Young (EY), $2.9 million.

5    At present, I have no real understanding of the extent of the substantive work undertaken by each respondent. There is no basis to assume that each party has progressed the matter at the same pace. It may be, for example, that EY has undertaken a considerable amount of “front-loaded” work, including by engaging consultants and the like, which has not yet been undertaken by other respondents. For reasons to which I will come, that does not presently matter.

B    SECURITY FOR COSTS

6    Although there was some argument today, the parties do not oppose orders being made requiring the applicants to provide $680,000 by way of further security for the future costs of each of Blue Sky and EY up to the completion of the conferral process put in place by earlier orders of the Court. This is more than the amounts ordered by consent in respect of Mr Shand and Mr Kain, but given the differing nature and extent of the case advanced against EY (and some additional work to be undertaken by Blue Sky), that is unsurprising.

7    Security for costs incurred prior to 27 August 2026 has been provided in respect of each of the other respondents in the amount of $570,000. EY seeks security for its past costs totalling $1.934 million, inclusive of the amount of $634,000 already ordered. Although the contrast is glaring, it is not perhaps entirely inexplicable. As I have said, the case against EY is broader than the case advanced against the other respondents.

8    In a case such as the present, it is only possible for the Court to take a fairly broad-brush approach to the provision of security, a fortiori in circumstances where, if the proceeding does not resolve following the conferral process and must be set down for trial, the Court will be required early next year to fix security for the balance of the proceeding. That is the course I have taken by implementing the interim provision of security today for a part of the interlocutory progress of the case.

9    There are, however, two matters which will bear directly upon the likely future cost of this proceeding and hence upon the appropriate amount of security which will ultimately be required which are worth stressing.

C    TWO ISSUES AND FUTURE CASE MANAGEMENT

10    The first concerns opinion evidence. There is a live question as to how the Court can best be assisted by such evidence in a proceeding of this scale and complexity. My present inclination is that, following the structured conferral process and the identification of granular questions upon which specialised opinion is genuinely required, close consideration should be given to whether some or all of those questions can be referred to an appropriately qualified referee to inquire and report. A “special jury” may provide a means by which specialised questions common to the cases advanced against several respondents can be examined coherently and economically, rather than through the traditional process of multiple experts preparing lengthy reports, responding to one another, conferring, producing joint reports and eventually giving concurrent or sequential evidence at trial.

11    I consider that a referee process has the potential to be substantially more cost-effective than receiving opinion evidence in the traditional form.

12    The appointment of a referee would not be novel in this proceeding. By way of example, a referee was previously appointed to inquire and report in relation to costs, and the resulting report was subsequently adopted: R&B Investments Pty Ltd (Trustee) v Blue Sky Alternative Investments Limited (in liq) (Adoption of Referee’s Report) [2024] FCA 1083.

13    I emphasise that I have reached no concluded view that a reference is the appropriate course, still less about the questions which might be referred or the identity of any referee. Those matters will require careful consideration after the conferral process has exposed what is genuinely in dispute. But this is precisely the stage at which the Court and the parties should be thinking inventively about the scope of the trial. There are highly experienced and able practitioners involved on all sides of this proceeding. At the appropriate time, I expect them to assist the Court in devising procedures which will allow the real controversies to be determined fairly and efficiently, rather than reproducing every expensive forensic practice simply because it has traditionally formed part of complex commercial litigation.

14    The second matter concerns, in broad terms, the likely extent of the documentary case at any trial. I recently addressed an aspect of this subject in the Sixteenth Sir Harry Gibbs Memorial Oration, All the Right Notes: Artificial Intelligence and the Future of the Common Law, delivered on 22 August 2026.

15    I there referred to commercial life having become documentary in a way earlier generations could scarcely have imagined and to the way in which the legal response to this abundance has largely concentrated upon improving the machinery by which documents are searched and reviewed. I described “teams of solicitors working like drones in light-deprived offices” becoming expert in managing electronic databases and observed that, notwithstanding technological progress, discovery and document analysis (and understanding a subset of the whole documentary corpus) has been hugely expensive in modern commercial litigation.

16    There is no shortage of primary information in this case; indeed, the problem is the opposite. There is an enormous quantity of contemporaneous material which will need to be collected, searched, understood and synthesised before it can be converted into a form useful to the parties and ultimately to the Court. Given no human being can sensibly comprehend such a corpus by reading it all, traditionally, solicitors create mechanisms by which the material is reduced: searches are devised, documents are coded, and with the assistance of counsel, chronologies are prepared and a much smaller subset of documents are selected for deployment at trial.

17    In the Oration, I suggested that artificial intelligence creates the possibility of interrogating the documentary corpus as a whole: to construct objectively generated chronologies from very large bodies of records; identify communications between relevant actors; trace the development of decisions; recognise documents which contradict an emerging chronology; identify apparent gaps in a sequence; and allow lawyers and judges to interrogate the documentary record by reference to propositions and issues rather than merely keywords.

18    The point is not to allow a machine to determine what happened. It is to use technology to undertake the immense work of organising and synthesising primary information so that the human beings charged with exercising professional and judicial judgment can identify what really matters. The importance of maintaining that distinction has also been noted in academic commentary: See also Professor Michael Legg, ‘Better than a Bot – Instilling Ethical Judgment into the Lawyers of the Future in the Age of AI’ (2024) 33(3) Griffith Law Review 273 (at 279–280). The abundance of information in modern complex litigation has created a problem which cannot sensibly be solved merely by adding more human beings to the task of processing it. If technology can perform some of that work reliably and at substantially less cost, it would be curious to insist upon the continuation of a labour-intensive process merely because it is familiar.

19    This proceeding seems to me an obvious candidate in which to investigate whether that possibility can be translated into practical case management. If the proceeding does not resolve, it will ultimately have to be tried as a formidable, documentary-heavy case. There is little sense in waiting until all those processes have occurred, and their costs have been incurred, before asking whether they could have been organised differently. The time to design a manageable trial is before the case becomes unmanageable.

20    This is not technological enthusiasm for its own sake. Courts have long emphasised the need to conduct litigation proportionately and to avoid unnecessary cost and delay. But those aspirations will achieve little if, when confronted with new means of performing entrenched and expensive tasks, courts and parties simply reproduce familiar processes because they are familiar. In litigation of this scale, proportionality cannot mean merely conducting conventional processes somewhat more efficiently. It requires consideration of whether the processes themselves remain appropriate.

21    The figures already expended in this proceeding illustrate the point. Millions of dollars have been spent by the respondents before any of them has filed lay or expert evidence. I do not say that those costs were unnecessary; as I have already explained, I presently lack the information which would allow me to form any such view. But the figures demonstrate the extraordinary cost which litigation of this kind is capable of generating before the parties have approached anything resembling a trial.

22    That has an obvious connexion with the applications presently before me. Security for the balance of the proceeding should not be assessed upon an unexamined assumption that every future stage will be conducted in the most expensive conventional manner. Before determining what security should be ordered for the balance of the proceeding, I want to know more about the structure of the case which the Court will actually permit to go forward. The amount reasonably required for future costs necessarily depends upon what reasonable future costs will be, and that in turn may depend substantially upon how the proceeding is managed.

23    Following the results of a conferral process being furnished to my chambers, I propose to examine with the parties how artificial intelligence and other appropriate technology might be harnessed in the circumstances of this proceeding to reduce further costs and to prepare and synthesise the vast body of contemporaneous documents which one can presently expect will provide much of the most reliable evidence relevant to the determination of liability. At the same time, the parties should be in a position to identify with greater precision the factual controversies which remain and the granular questions upon which specialised opinion evidence is genuinely required.

24    At that point, I will expect the parties to address two related questions: the first is whether identified questions of specialised opinion can more efficiently be dealt with through one or more appropriately qualified referees; the second is how technology, including artificial intelligence, can appropriately be used in documentary synthesis and preparation for trial. These questions about the mechanics of litigation will determine whether the case can be conducted in a way which is proportionate.

25    I am deliberately not prescribing how the latter task is to be performed, and I do not assume that any particular commercial product or technological method will provide the answer. There are sensible people involved in this proceeding, with access to substantial professional and technological resources. I expect them to obtain whatever specialist assistance is required and to approach the problem cooperatively and with some imagination.

26    Any methodology ultimately adopted would plainly require appropriate safeguards. It would have to preserve the capacity to trace propositions back to the underlying primary documents, permit the parties to identify error and challenge or supplement any resulting synthesis, protect privileged and confidential material where necessary and remain subject to human supervision. The parties must remain free to adduce and rely upon the evidence by which they seek to prove their respective cases. But none of those requirements provides a reason for declining to investigate whether technology can provide an objective and useful starting point.

27    The proposed referee process, and the possible use of artificial intelligence should not be regarded as unrelated experiments. They reflect the same approach to case management. Highly skilled lawyers and judges should be spending their time upon tasks which require professional and judicial judgment. Where specialised questions can be isolated and examined more efficiently by an appropriately qualified referee, that possibility should be considered. Where the collection and synthesis of enormous quantities of primary information can be performed more efficiently by technology, that possibility should equally be investigated.

28    In the end, the Court must be placed in a position to understand an enormous body of primary material without the trial becoming an exercise in documentary attrition; the parties must retain the capacity fairly to advance the cases they wish to make; and this must occur at a cost proportionate to the controversies requiring determination. I propose, with the assistance of the parties, to do the work necessary to achieve that result.

29    Only after those matters have been addressed will I be in a sensible position to consider the amount of security which should be ordered for the balance of the proceeding.

D    SECURITY FOR EY’S PAST COSTS

30    That leaves the outstanding issue concerning security for the past costs incurred by EY. With her customary vigour, Ms Morgan SC urged upon me that a very large amount of security should now be ordered. Despite those entreaties, I have decided, at least at this stage and until I come to determine security for the balance of the proceeding, to maintain some degree of proportionality between the respondents.

31    Senior counsel for the applicants informed me that the applicants are prepared to offer an additional $600,000 by way of security, in addition to the $634,000 already ordered. I consider that to be an adequate interim position.

32    For completeness, when the Court comes to determine security through to the conclusion of the trial, I do not regard the order made today as foreclosing any submission by EY that it has undertaken more work to date than the other respondents and that this should be reflected in the amount of further security ordered in its favour. That question is for another day.

33    Accordingly, I direct the parties to bring in short minutes of order giving effect to these reasons and the agreements they have reached.

Associate:

Dated: 1 September 2026

SCHEDULE OF PARTIES