Chief Executive of the Department of Corrections v Fujitsu New Zealand Limited

[2023] NZHC 3598

IN THE HIGH COURT OF NEW ZEALAND WELLINGTON REGISTRY

I TE KŌTI MATUA O AOTEAROA TE WHANGANUI-A-TARA ROHE

CIV-2021-485-423

[2023] NZHC 3598

BETWEEN

HIS MAJESTY THE KING IN RIGHT OF NEW ZEALAND ACTING BY AND
THROUGH THE CHIEF EXECUTIVE OF THE DEPARTMENT OF CORRECTIONS
Plaintiff

AND

FUJITSU NEW ZEALAND LIMITED

Defendant

AND

DASSAULT SYSTÈMES AUSTRALIA PTY LIMITED

Third Party

Hearing:

11 September – 5 October 2023

Appearances:

M G Colson KC, K J Dobbs and M R M Gale for Plaintiff

C L Elliott KC, M B Wigley and J Kohu-Morris for Defendant
C F Finlayson KC, A J Horne, H M Jaques and C Hoeft for Third Party

Judgment:

8 December 2023

JUDGMENT OF COOKE J

Table of Contents

Factual background  [5]

The Registrations of Interest phase  [7]

Provision of information  [17]

Fujitsu’s RFP response  [23]

Demonstrations of Quintiq  [38]

Further information  [46]

Dassault’s concerns  [53]

SOW23  [66]

Revised pricing  [71]

Gap between out of the box and required functionality  [83]

CHIEF EXECUTIVE OF THE DEPARTMENT OF CORRECTIONS v FUJITSU NEW ZEALAND LIMITED [2023] NZHC 3598 [8 December 2023]

Problems develop  [97]

What went wrong?  [105]

Breach of contractual warranties  [113]

The warranties  [114]

The nature of warranties  [120]

What was warranted?  [124]

Customisation v configuration  [132]

Out of the box  [136]

Seamless integration  [139]

The price  [143]

Other matters  [146]

Representations during negotiations  [149]

Warranties qualified  [151]

Were the representations untrue?  [155]

Out of the box with no customisation  [156]

SAP integration  [158]

Price  [162]

Mr Driessen’s analysis  [165]

Conclusion  [173]

Department’s other claims against Fujitsu  [174]

Additional contractual obligations  [175]

Misrepresentation  [182]

Fair Trading Act  [191]

Criticisms of the Department  [197]

Loss  [198]

Assumption of responsibility  [201]

Exclusion clause  [205]

SOW23 costs  [208]

Third party costs  [209]

Admissibility of new spreadsheet  [214]

Has the Department proved the claimed loss?  [219]

Work reusable  [232]

Conclusion  [240]

Fujitsu’s claims against Dassault  [242]

Does the CCA apply?  [244]

Claims under the FTA  [254]

Liability under the CCLA  [265]

Exclusion of FTA liability  [276]

Interpretation issues  [277]

Application of s 5D  [282]

Destruction of documents  [290]

Conclusion  [297]

[1]    The Department of Corrections (the Department) manages a significant workforce in order to operate the prison system over approximately 18 prisons throughout New Zealand. In December 2018 it entered  a  contract  with  Fujitsu New Zealand Ltd (Fujitsu) under which Fujitsu was to provide new software for the Department to manage the rostering of its staff in a more efficient way. Fujitsu’s proposal involved software provided by its sub-contractor, Dassault Systèmes Australia Pty Ltd (Dassault) and a Dassault product called “Quintiq”. The Department purchased the licence for the Quintiq software in December at a cost of $1.8 million. The contract with Fujitsu involved the analysis and design phase for the implementation of Quintiq as the rostering solution. The contracts were entered following an earlier Request For Proposal (RFP) issued by the Department in March 2018. In its RFP response Fujitsu had stated that the Department’s requirements could be met “out of the box” by Quintiq without the need for customisation, that it could be implemented seamlessly with the Department’s existing payroll systems in accordance with the Department’s timeframes, and for the approximate total cost of

$716,000 over and above the licence costs.

[2]    In June 2019, following a period of substantial work by all parties under the analysis and design phase, the Department brought this contractual arrangement with Fujitsu to an end. This followed Fujitsu supplying a revised pricing proposal that put the cost of the total project at closer to $7 million in addition to the licence cost. The Department subsequently contracted with another company to provide a rostering solution.

[3] The Department now sues Fujitsu for breach of contractual warranty and other contractual terms, for misrepresentation under the Contract and Commercial Law Act 2017, and for misleading and/or deceptive conduct under the Fair Trading Act 1986. It says that Fujitsu’s warranties and representations about Quintiq were untrue. Fujitsu in turn sues Dassault under the Fair Trading Act and the equivalent Australian legislation (the Competition and Consumer Act 2010), and for misrepresentation under the Contract and Commercial Law Act, for any liability arising by virtue of the Department’s claims on the basis that any untrue or misleading statements about the Quintiq product originated from Dassault.

[4]    The Department sues Fujitsu for a total of approximately $4.3 million for the wasted expenditure it says it incurred on the project. This involves approximately

$640,000 that it paid Fujitsu for work before the contract was ended, $1.8 million which it paid for the Quintiq licence, and approximately $1.9 million for other costs that the Department says it incurred as part of the project.

Factual background

[5]    I begin by outlining the facts. This will include making factual findings, although it will also be necessary to make additional findings when addressing the particular claims.

[6]    I observe from the outset that the primary source for making findings is the contemporaneous documentary record. Whilst some of the oral evidence has been of assistance I found the contemporaneous records to be the most reliable source of evidence. The events occurred some years ago, before the COVID-19 pandemic, and I generally considered that much of the oral evidence involved an attempted reconstruction of events, albeit based on the contemporaneous documents, rather than true recollection.

The Registrations of Interest phase

[7]    The new rostering system was intended to be implemented by the Department as part of a wider project for reforming the Department’s approach to the rostering of prison staff which it initiated in 2016. This was called the “Making Shifts Work” project. An important aspect of this project was a desire to allocate staff more efficiently, and in a way that it was hoped would improve the rehabilitation of prisoners. Steps were taken in late 2017 to identify firms who might be able to provide a new software for rostering as part of that project.

[8]    Fujitsu was already a contractual partner with the Department. Pursuant to a Master Services Agreement dated 22 December 2015 (the MSA) a contractual relationship had been established. The MSA had been entered so that Fujitsu could provide development and maintenance support associated with the use of applications by Department staff (such as those associated with mobile devices and Microsoft

products) but the MSA contemplated that other services could be brought within the terms of the MSA.

[9]    On 10 November 2017 the Department called for Registrations of Interest (RoI) for the rostering solution through the New Zealand Government’s Electronic Tender Service. The RoI explained that the Department needed a new rostering system for 6,000 staff across 18 operational prison facilities and other sites which changed its current shift patterns so it could more effectively operate the prison system in a manner that would better rehabilitate offenders. In it it said:

What we do not want

The Department is not looking for a solution that is onerous to the business, places current operations under strain, and does not integrate with existing systems. The aim here is for the Department to implement a solution that increases operational efficiency, is flexible, scalable and can be adapted further in future.

[10]It also said:

Integration Requirements

Any rostering solution should seamlessly integrate with our current SAP Payroll system. Considering the complex requirements of shift patterns and significant data exchange occurring between systems, it is critical that any solution can facilitate improved operational performance rather than being a burden on the business.

We are interested to hear how Respondents interface patterns can integrate with existing systems to enable a more integrated end-to-end solution.

[11]   The SAP system was used for the Department’s payroll. So the rostering solution needed to integrate with SAP, which was a commonly used payroll system, and deal with the complex requirements the Department said were involved.

[12]   Dassault saw the RoI and gave its own consideration to seeking to contract directly for the project. But its assessment was that it appeared a small project, and that it would be better to engage in the work as a sub-contractor to Fujitsu who had an existing relationship with the Department. Dassault and Fujitsu had been involved in other tender bids in New Zealand. For its part Fujitsu also wanted to be the contractor with Dassault as its sub-contractor.

[13]   Even at this early stage Dassault foresaw that the project would only be a profitable one for it if there was greater complexity. In emails in November 2017 from Ms Kate Gayner (a manager at Dassault who subsequently supervised the pre-sales team) to Mr Nigel Deans (who became the lead-manager for the project) and others Ms Gayner advised that what was being sought was “cheap” and “basic” and she asked “how are we going to extend the complexity to make this worthwhile responding to?” In any event Dassault and Fujitsu agreed that Fujitsu would respond to the RoI, and the subsequent RFP, and Dassault would be Fujitsu’s sub-contractor.

[14]   The Department received eight responses to the RoI. Fujitsu’s response was dated 29 November 2017. Under its response Fujitsu would be the Department’s contracting party, and Fujitsu would sub-contract Dassault to provide Dassault’s Quintiq product, which came to be known as the “Quintiq solution”. On 13 March 2018 Fujitsu and Dassault entered a “Teaming Agreement” to record the terms of their arrangement.1

[15]   Fujitsu said to the Department in the RoI response that contracting with it using the Quintiq product supplied by Dassault would mean that it could “deliver an off-the- shelf, integrated solution that can support all of [the] Department[’s] requirements”. It also said that amongst the key points were that the solution could meet the Department’s “core functionality out of the box” with “minimal requirements [for] customisation” and “seamless integration” with the Department’s SAP payroll system. At trial there was debate about what some of these expressions meant. I will address this debate in greater detail below. But in essence these statements conveyed that the Quintiq software could easily integrate with the Department’s existing SAP payroll system in order to provide the Department’s core requirements, and that limited changes would need to be made to the Quintiq standard product to do this.

[16]   The Department was using the services of the accounting and consultancy firm Deloitte throughout the wider Making Shifts Work project. Deliotte evaluated the RoI responses on the Department’s behalf, and Fujitsu was assessed as being one of the

1      Fujitsu and Dassault later signed another agreement dated 10 April, called a “System Integrator Alliance Agreement” which also appears to regulate the agreement between them, but which was not specific to this project.

top four potential vendors. Deliotte then recommended, and the Department approved, a “lean” RFP process. Under this approach the Department would not have fully documented business requirements which it would issue to the market as part of the RFP, but it would describe its requirements in more general terms, and then work together with the proposed vendors it would choose through the RFP process. The expert witness called by Dassault, Dr Kenneth Tan, was critical of this approach. In his view it almost guaranteed that there would be changes in scope and increases in cost. He referred to examples of such projects he had been involved in in Australia which had ended badly. But he accepted that he had no experience of the New Zealand market. I accept the evidence of the other witnesses, including the expert evidence from Mr Mark Peach called by the Department, that a lean RFP is a standard approach in New Zealand and that it does not necessarily lead to greater cost in delivering a solution.

Provision of information

[17]   Two vendors, including Fujitsu, were identified by the Department as the top contenders after evaluation of the RoI responses. On 18 January 2018 the Department advised Fujitsu that it had been short-listed, and was invited to proceed to the next stage. On 9 February the Department’s sent 30 “use cases” to Fujitsu as a part of the lean RFP process. These were 30 scenarios where Fujitsu was invited to demonstrate how the Quintiq product would be used to address the scenarios. On 26 February the Department then issued its Solution Requirements, and the RFP itself was released on 1 March 2018.

[18]   It is important to record an issue about the Department’s requirements at this stage. As had been indicated in the RoI there were complications with the way the Department undertook rostering. The roster had more simple requirements — prison officers being scheduled to work shifts reporting to senior officers within a particular prison. But there were also complexities. Such officers could also work some of their time at other prisons in the same region reporting to different senior officers. Moreover within each prison there were separate “units”, such as a high security unit,

or a part of the prison used only for remand prisoners.2 Units needed to be able to be separately managed. Prison officers could also be required to move prisoners between prisons, or to participate in other tasks such as bringing prisoners to court for hearings. This kind of activity could also be allocated to different “cost centres” — that is, some staff activities needed to be managed and recorded to different budgets. Sometimes officers would be allocated tasks normally performed by more senior officers, and entitled to higher pay. Information needed to be sent to the payroll system to record such activities. For example prison officers might be required to take a prisoner currently on remand in a prison van to a court for a hearing, and then transport that prisoner back at the end of the day to a different prison to a unit for sentenced prisoners. Such a scenario might involve the prison officers being under the authority of different more senior prison officers at different times, and their activities required to be allocated to different cost centres and/or at higher rates. This might all occur during a single shift for the prison officer. These kind of features created more complex rostering requirements.

[19]   This means that the information provided about the Department’s operations, and what was said about Quintiq’s capabilities is important. As indicated, the RoI referred to the “complex requirements of shift patterns and significant data exchange occurring between systems”. On 28 February, two days after the Department had released its Solution Requirements Fujitsu/Dassault then attended a  meeting  in  New Zealand to gather more information about these requirements. Ms Nicola Horwood the manager at Fujitsu who had lead responsibility for the tender made a note of the meeting. There was no other substantial note made of the meeting available from Dassault  or  the  Department,  although  Ms Horwood  sent  her  note  to Messrs Deans  and Moran  of Dassault for comment (with no response received).  Ms Horwood’s note confirms the evidence that some of the complexities of the Department’s rostering requirements were either identified, or greater potential complexity was at least foreshadowed when the Department explained its operations. This included the following explanations:

2      Remand prisoners are those in custody yet to face trial who are kept separate from sentenced prisoners and subject to different conditions (such as limited rehabilitation programmes).

(a)That the Department’s roster requirements operated at the prison level, but it could also have regional rostering requirements. For example in the Auckland region staff could move around prisons in the same region under the control of different senior officers. In addition staff could be moved across regions.

(b)Each of the prisons had a number of separate units which involved different compositions of officers. There were three main categories of officer, a Principal Corrections Officer (PCO), a Senior Corrections Officer (SCO) and a Corrections Officer (CO). Even when an officer worked in a single unit for an eight hour shift they could do various tasks in that unit which needed to be allocated to different cost centres in the Department’s system. The Department also wanted the capability to “cross-charge” costs to other units.

(c)There were various other rostering complexities, including an officer being paid at a higher rate when fulfilling a more senior role (that is, a CO doing the job of a SCO), and situations where shifts were “gifted” from one officer to another.

[20]   Mr Moran of Dassault attended this meeting. He subsequently had a significant role in demonstrating the Quintiq product to the Department representatives at subsequent meetings in New Zealand. His role was essentially in sales. He attended to obtain information for the purpose of these demonstrations. He gave evidence that the information received at this meeting was of some assistance for his demonstrations of Quintiq, but that it “went beyond the level of detail that would typically be incorporated into a product demonstration in the sales cycle”.

[21]   The Department’s Roster Solution Requirements issued on 26 February did not include details of all the complications discussed at the meeting two days later. For example the Requirements did not identify the need to allocate activities to cost centres based on particular tasks. But the purpose of the meeting was to provide further information, and I accept that Fujitsu/Dassault were put on notice of the potential complexities. For example, the Requirements referred to there being different cost

centres associated with the rostering, and it had been explained at the meeting that the tasks officers were to undertake on particular shifts were to be allocated to different cost centres.

[22]   Various other steps were taken before the RFP responses were required. In addition to the release of the Roster Solution Reports on 26 February, and the meeting on 28 February, Dassault conducted its first product demonstration on 7 March, and there was an implementation workshop on 9 March. The Department also issued clarifications on 5 March to enable all tenderers to obtain further information about the Department’s requirements. These included clarifications issued to  Fujitsu on  14 March.

Fujitsu’s RFP response

[23]   Fujitsu submitted its RFP response to the Department on 16 March. This document is of central significance to the Department’s claims as it is largely the basis for the Department’s claims of misrepresentation and breach of contractual warranties, and it is also a key aspect of the alleged misleading and deceptive conduct. In addition Fujitsu’s claim against Dassault is based upon the information Dassault provided to Fujitsu for the purposes of this RFP response.

[24]   The document is lengthy and includes a number of statements about the Quintiq solution, including by addressing the listed requirements set out in the document issued by the Department on 26 February. I accept the arguments for Fujitsu and Dassault that the RFP response needs to be read as a whole, and that it is important not to read individual statements without understanding the full context of what was being said about the Quintiq solution, including in the context of more precise statements in the main body of the document. But a key element of the claims advanced by the Department arises from statements made in the executive summary. This stated:

Summary

The Department can be confident that in selecting your existing partner Fujitsu, and our advanced technology partner Quintiq, you will be provided with a solution that is designed to:

·     meet all of the Department’s core functionality requirements;

·     require no customisation and is ‘out of the box’;

·     be delivered with a trouble free implementation;

·     meet the Department’s tight timeframe for delivery of a fully-functional solution that is ready for internal testing by 15 February 2019;

·     integrate well with SAP;

…

[25]   Earlier in that summary Fujitsu also stated that the Quintiq solution would involve “seamless implementation”.

[26]   The RFP response also addressed 97 more detailed requirements. These were broken into functional requirements, and non-functional requirements, identified those which were “must have” requirements and those that were not, and also cross- referenced the use cases that had been issued. The response indicated that Fujitsu could meet the listed functional and non-functional requirements in full with the exception of three non-functional requirements which it indicated it “partially met”. For example functional requirement number 11 contemplated information being provided to the SAP payroll system involving a number of details, including that the costs of an allocated work could be allocated to a different unit, and an ability to make changes to this retrospectively. When responding that this requirement was “fully met” the response stated:

Quintiq has a wealth of experience in integrating Time & Attendance data from the rostering system to payroll systems such as SAP. A number of those experiences involving integrating with SAP Payroll (via Substitutions and Attendances) include the fully costed pay data. Additionally, Quintiq Time & Attendance [provides] a means to retrospectively change time entries. Some configuration may be required depending on the payroll period and period in which changes can be made.

[27]   The last sentence is the kind of more particular statement that is a reflection of what was said in the executive summary. Some configuration (or changes) might have been needed to allow Quintiq to make retrospective changes to time entries if the changes related to longer periods of time, or needed to be made at some later date. But otherwise the product could meet the Department’s needs “out of the box” — that is, it could be met by Quintiq’s standard functionality.

[28]   There are other similar statements in relation to the more particular requirements. Read as a whole I accept that the executive summary was an accurate summary of what the body of the document also represented about the Quintiq solution. Put another way, a reader of the document would not understand the subsequent more particular content provided in relation to each of the functional and non-functional requirements to materially qualify what was said in the executive summary in a substantive way. The only qualifications, or assumptions, were those that were expressly set out in the RFP response. I address those below.

[29]   Representations were also made in relation to the price. For that purpose the response indicated what the Department’s core requirements were by reference to the following table:

[30]   The price provided was for “release 1”, being the core requirements, encompassing those tasks shaded in green in the middle of this table and summarised on the right hand side. The pricing schedule provided with the response identified a price on a time and attendance basis estimated at $716,000.

[31]   As indicated the RFP response also set out assumptions, including those associated with this price. It was said to remain valid up until 26 September 2018, and that it was based on the “high level” requirements provided by the Department with the final scope of the project to be agreed as part of the statement of work.

[32]Other assumptions were expressly set out including that:

All business requirements, process mapping and rostering rules have been documented prior to commencement of the project.

[33]   It is of significance that the RFP response did not state that what Fujitsu said about Quintiq depended on the business requirements not being more complicated than what Fujitsu/Dassault assumed for the purpose of the RFP response. Neither did it include any other qualification of that kind — for example an assumption that the Department sought only Quintiq’s out of the box functionality. Rather, it proceeded on the basis that all the Department’s requirements could be met by the Quintiq solution. The only relevant assumption was that the detailed business requirements would be provided before the work started — it was only an assumption relating to the timing of their provision. The RFP had expressly required that responses included “all assumptions and qualifications made about delivery of the Requirements”. Fujitsu’s RFP response also stated that it was provided based on “all … information provided by the Department of Corrections … to Fujitsu regarding the Rostering Programme 2017 requirements” which includes the information provided at additional meetings as well as the formal information provided in the Rostering Solution Requirements.

[34]   There was, however, one significant and overriding qualification in the RFP response. It was made clear that the response was not an offer capable of acceptance, that it was based on what the Department had disclosed, and only so that the Department could form a view on whether Fujitsu would be invited to participate further in the process. It stated:

In the event that the Department amends its requirements in the Rostering Programme 2017 or otherwise as they apply to the Response, Fujitsu reserves the right to amend the Response accordingly, including scope, pricing and other related information or requirements.

Any pricing that has been submitted by Fujitsu is only indicative. Fujitsu will provide to the Department clear and binding pricing and other detail once Fujitsu responds to a more specific document containing the relevant information needed to provide such pricing.

[35]   Given this qualification and other statements to the same effect much depends on what happened after the RFP response was sent and the contracts were later entered.

Even if the RFP response contained misrepresentations — which for reasons I address below I consider that it did — these might only give rise to legal liability because of what happened after the RFP response was sent. That is because it was expressly premised on Fujitsu’s ability to confirm the position for the purposes of entering contractual arrangements at a subsequent point.

[36]   The RFP response was evaluated by a panel selected by the Department. Clarifications were sought by that panel in March and April. Reference checks were also undertaken. A due diligence report was provided in June leading up to Fujitsu being appointed preferred vendor in July.

[37]   I accept the point emphasised by both Fujitsu and Dassault that the Department’s requirements as provided in the RFP were very high level. The response to those high level requirements also made it clear that the legally binding obligations could change with further information. I also see force in Dassault’s point that well- resourced and sophisticated commercial parties cannot reasonably rely on general impressions that may be given in pre-contractual documents. But these points only take Fujitsu and Dassault so far. The nature of their response to the high level requirements nevertheless included a clear representation — that Fujitsu/Dassault had an out of the box solution which could meet these requirements without the need for significant changes to it. That representation was not qualified — for example by saying it could do so provided that the Department’s detailed business procedures did not reveal complexities. So the high level request for a proposal was met by a high level representation of what Fujitsu/Dassault could provide. If Fujitsu/Dassault were to qualify that representation they needed to do so squarely before any contracts were entered. For the reasons addressed below I consider that the representations were confirmed rather than qualified by Fujitsu and Dassault’s statements and conduct through to the entry of the contracts, including the main contracts entered in December 2018.

Demonstrations of Quintiq

[38]   A key aspect of the reiteration of the statements made in the RFP arose from the product demonstrations given by Dassault.  These were primarily provided by  Mr Moran. The first of these presentations was on 7 March shortly before Fujitsu’s

RFP response, and there was a further presentation on 20–21 August at a roster technology workshop after the RFP response.

[39]   Evidence was given by the Dassault personnel involved in the demonstrations, the Departmental personnel to whom they presented, and the Fujitsu personnel who were also present. These demonstrations were either undertaken “live” — where the Quintiq product would be displayed and the rostering tasks it could undertake demonstrated — or they were pre-recorded and then spoken to at a presentation. One of the recordings of the demonstrations was produced in evidence, and the Court was provided with a series of screenshots from it as well as having the presentation played.

[40]   The demonstrations were intended to show the Quintiq product in operation. They involved a projection of a computer screen showing how the Quintiq rostering solution would be portrayed to a user making rostering decisions. The roster was presented in the form of a table on the screen. The table would have prison sites/units down the left hand side and staffing requirements for time periods across the top. The user could drag and click listed employees into the roster for the site/unit to meet the rostering requirements. This included allocating employees to particular shifts at particular places for particular times. It took into account complications such as split- shifts and when an employee was working in a higher duty capacity. The demonstration also dealt with more complex tasks, such as the allocation of equipment to the employees for those tasks, again by a drag and click process.

[41]   These demonstrations were very effective from a sales perspective. Mr Moran said he took pride in his demonstrations, and I consider that they played a significant role in the Department confirming that it would contract with Fujitsu for the supply of Dassault’s product. I generally accept the evidence of the Department witnesses, which was not significantly disputed by the Dassault witnesses and confirmed by the Fujitsu witnesses, that the demonstrations suggested that the Quintiq solution could meet the Department’s requirements with its existing functionality, and that it could integrate with the Department’s existing payroll system. It portrayed the Quintiq solution as a highly attractive product that would meet the Department’s needs. It is apparent that the Department (and indeed Fujitsu) was very impressed by the demonstrations.

[42]   The evidence also establishes, however, that the Quintiq product as presented had been worked on by Dassault for the purpose of the demonstrations and did not simply involve its standard (or “out of the box”) functionality. Mr Moran gave evidence that for ten of the use cases he was demonstrating Quintiq needed to be configured so that it could address what was required. He explained that doing such work on the Quintiq product prior to the demonstration cost time and money. So he had addressed the nature and extent of this work with Ms Gayner. She supervised the pre-sales team, including Mr Moran. They dealt with this issue before the demonstration in March. They agreed to Mr Moran making changes so that the Quintiq product could deal with features that could not be met with out of the box functionality. An example was the higher duties allowance — when an officer was to be paid more for acting in a more senior role. This was addressed by entering additional data into the demonstration as it was a feature that the standard functionality did not address.  In her internal communication with Mr Moran about this change  Ms Gayner referred to this as doing something “dodgy with the data”. She said in cross-examination that this was simply a poor use of words. Other more complex tasks also required the demonstration to be altered as the standard Quintiq functionality could not achieve the tasks.

[43]   Mr Moran said under cross-examination that he would have told the Department’s representatives that the Quintiq product had been configured for the purposes of the demonstrations, and that he was not simply demonstrating standard functionality. That is not consistent with the evidence of the Department witnesses and I do not accept that he provided a qualification in those terms. I agree that it would have been apparent that information had been entered into Quintiq in order to produce the demonstration — for example, the demonstrations listed particular New Zealand prisons and involved the Department’s categories of prison officer. So the demonstration was obviously not the Quintiq product without any customer specific inputs — it was apparent that information would have been entered into it in order that Dassault would demonstrate how it would work for the Department. But it was nevertheless reasonably understood to be portraying Quintiq’s standard functionality.

[44]   Mr Kathiresan, Fujitsu’s project manager, attended the presentation on 20–21 August. He gave evidence that everyone was amazed by how Quintiq performed, and

that he understood that what was being demonstrated was its “out of the box” functionality. He also said that the fact that this turned out not to be so was a major issue as it fundamentally changed the principles that the parties were working on. I accept his evidence in this respect. If Mr Moran did say anything of the kind he suggested in cross-examination it was not, and would not reasonably have been understood to be a qualification upon the attributes of Quintiq that he was demonstrating. Mr Moran struck me as a polished and effective salesman, and any comment he made did not qualify, or diminish the capabilities of the standard Quintiq solution that he had the goal of selling. Mr Moran also explained that the purpose of the presentations, including those on 20 and 21 August was to get the Department’s team “… excited about upcoming changes to their rostering, as this would help the Programme to be a success” — that is, to achieve the sale. This accurately captures Dassault’s approach to the demonstrations.

[45]   I accept the Department’s argument that the demonstrations reiterated the RFP representations, and they then suggested that the Quintiq standard functionality could do the tasks shown at the demonstrations (such as addressing the high duties allowance) with standard functionality.

Further information

[46]   Further information was provided to the Department in the RFP process which also confirmed the representations.

[47]   On 28 March the Department issued a list of clarification questions which included, in relation to one of the non-functional requirements, whether “during discussions and review of our Requirements, did you note the need for any customisations?” The response dated 3 April stated:

During discussions and review of your Requirements, we have not identified the need for any customisations.

Fujitsu has provided a solution that leverages out of the box product capability. In certain cases, configuration may be required to meet specific requirements.

[48]   This effectively repeated what was said in the RFP response. It conveyed that changes to the Quintiq product might be required in relation to particular matters, but

what the Department was being provided was a solution based on Quintiq’s standard functionality. The reference to “discussions” made it apparent that this assessment was based on what had been addressed at meetings with the Department and was not limited to the formal documentation.

[49]   Customisation was something the Department wished to avoid. In July 2018 the Department’s Technical Advisory Council approved the Reference Architecture for adoption of the Quintiq solution. In doing so it recorded “customisations in a sense of system modification must be avoided at all costs and some customisations might need to be approved …”. This is clumsily worded, but meant that customisation was to be avoided, and may need to be approved by this body if proposed.

[50]   There were two related reasons why the Department wished to avoid customisation. The first was that it was likely to be more expensive — the more changes that are made to a standard product the more work and time it would take. Secondly such changes potentially compromised the ability to upgrade the product in the future. The Department wished to stay with the standard product so that it would remain upgradeable.

[51]   I accept that through the further processes the Department became aware that changes would be needed for implementing the Quintiq solution particularly for integrating the solution with the Department’s SAP system before it entered the contracts in December 2018, however. Integration with SAP was likely to be more complex and expensive than Fujitsu had portrayed in the RFP response. This became a material qualification on what Fujitsu had represented. I address this further below. But this was the exception. The Department otherwise understood that the Quintiq solution could be implemented without significant adjustments in the way that had been portrayed in the RFP response. The Quintiq solution would remain a standard product which was upgradeable, and which avoided the expense of a customised solution.

[52]   Between April and June 2018 the Department undertook the reference checks, which were with certain New Zealand or Australian companies that had used the Quintiq product. There was no focus on the extent of any changes required to the

standard product in these checks. After doing so, however, Fujitsu was appointed as the preferred vendor on 9 July 2018. The parties then had an initial scoping workshop on 24 July. Initial drafts of the contract for the analysis and design phase were then prepared and exchanged.

Dassault’s concerns

[53]   During this period Ms Gayner began to look more closely at what had been stated in the RFP response, and she became concerned. On 14 May she had an email exchange with Mr Deans and Mr Moran to confirm that she understood that what was being provided in stage one was for “basic rostering only and implementing our out of the box solution” for the price indicated. But she had checked the RFP response and she realised that what was represented was not just basic rostering but also full pay code calculations — that is that it transmitted additional information to allow SAP to make the required payments — and that “that’s going to be a major problem”. This was the first indication of concern that the RFP response had misrepresented the standard functionality of Quintiq being provided for the indicated price.

[54]   She then looked at the RFP further. On 15 July she then sent Messrs Deans and Moran a further email stating:

I am feeling a little sick looking at the NZ Corrections tender. Based on the latest tender response there is no functional scope assumptions linked to our estimates. At this stage we will be on the hook to deliver everything for the pricing we have submitted.

The timeline of 4 weeks analysis, 4 weeks modelling and 4 weeks ID looks fine for a rostering out of the box solution, but this also includes T&A, which may cause a lot of problems.

I remember at the time pushing for a list of functional assumptions and I was told these would be completed and added to the RFP but I can’t see them anywhere. Our only god-send is Nicola added the assumption from the MPI tender stating we are only integrating to HR and Payroll as part of the first release. Just so you know the SOW is stating they want integration to BI/Finance and a whole lot of other systems in the first release.

There are also no assumptions around reporting/audit trail etc and their RFP is full of requirements around this. We need to find out who wrote this tender response as [there] has been no pricing assumptions in the individual response, just that we do everything basically.

[55]   The reference to Nicola is to Nicola Horwood, and to a change to the RFP using words from another tender document involving the Ministry of Primary Industries, so that only more basic integration with SAP was required at stage one rather than full pay code calculations as proposed in the draft statement of work (although the time and attendance issue was foreseen by Ms Gayner as still likely to later cause a major problem). Ms Gayner’s focus at this time was based on the price of the stage one functionality being provided — that what was being stated in the RFP response could address the core requirements described in the RFP at the price indicated. The “auditing” and “reporting” functionalities were in addition to the “pay code calculations” functionality she had initially raised. Her concern was that the Department’s stage one needs would involve more than out of the box functionality, that Dassault could not provide this for the indicated price, and in addition that further work on the time and attendance issue would reveal “a lot of problems”.

[56]   Mr Deans responded to the email with the suggestion that they should “manage our way through it”. I am satisfied that this is what Dassault decided to do. This approach involved seeking to reduce the functionality being provided to the Department for the stage one and adding it back in later as additional functionality at a greater price. For example on 19 July Ms Gayner made changes to the draft contractual documents — the Statements of Work — for the project “to try and pull back scope”. This included identifying elements that were said to be out of scope.

[57]   These email exchanges involved a recognition by Dassault that the RFP response had misrepresented Quintiq’s standard functionality. Even the stage one functionality would require work and was not “out of the box”. Dassault decided not to tell either the Department or Fujitsu that the RFP response had misrepresented Quintiq’s out of the box capabilities. While Ms Gayner’s focus was on what Dassault was providing for the stage one indicated pricing, the underlying issue was that the out of the box/standard functionality did not meet the needs the Department had outlined and that problems would be revealed once more detailed work was undertaken. I also accept that Dassault decided to address this problem by seeking to increase the price that would be sought on the basis that the Department had changed the scope of the project to encompass more than originally indicated, indicating that the Department was responsible for the cost increases.

[58]   Further information was then provided by the Department to Fujitsu/Dassault. On 13 August 2018 the Department provided the Making Shifts Work Reference Architecture. In addition the Department’s level one processes and business requirements were sent at the same time. This included further identifying some of the complications with the Department’s requirements, including that officers could work across different units with different cost centres as well as the potential for different reporting officers, and complications with leave or overtime requirements.

[59]   On 16 August 2018 Ms Stewart then sent Ms Gayner an email concerning the Department’s organisational structure and how it would fit into Quintiq’s unit structure which she said was “a gnarly requirement” and a “possible challenge”. Ms Gayner responded by indicating that it could fit into Quintiq “automatically” although a configuration change would be needed if more than one person had to sign off a rostering decision. This response again effectively reiterated the RFP response that such matters could be addressed by Quintiq’s standard functionality with configuration only needed for particular issues.

[60]   In closing submissions counsel for Fujitsu emphasised a further exchange of information in which the Department had asked particular questions, and Dassault had provided answers. On 28 August 2018 the Department asked a list of questions including:

For the customisation of the product, what tools are required to customise the product? OR are there tools bundled into the product itself. If so what are they?

[61]In its answer Dassault said:

Quintiq is proposing that zero customisation be made to the Quintiq application. We are proposing that configuration shall be required of the Workforce Planner industry solution. Regarding configuration, Quintiq provides four levels of configuration …

[62]   The four levels of configuration then described matters that could be regarded as significant changes, and as “customisation” in the eyes of those with technical knowledge. I consider that this answer does not greatly assist Fujitsu and Dassault, however. The answer expressly stated that there was no customisation proposed for the Department. Whilst it also said that there would be some configuration, there is

nothing in the answer that identified for the Department that significant configuration, or significant cost, was involved in providing Quintiq to the Department. That is because none of the examples of more significant “configuration” were linked to work to be done for the Department.

[63]   It is also apparent that a policy decision had been made by Dassault when dealing with potential customers to use the word “customisation” in a particular way. The Quintiq solution involved three layers — the base layer, a second layer involving a Dassault standard product (in this case a product called Workforce Planner), and a third layer which involved integration with the customer. It reserved the word “customisation” for changes to the base layer. It is also apparent that Dassault never made changes to its base layer, or may only have done so on one occasion. It had a policy of referring to changes to the layers above the base layer as “configuration”, even if there were substantial changes. That was so even if there were changes to its standard product — here Workforce Planner — in the second layer. By adopting this policy it was able to say that its products involved no customisation at all — the representation it made and repeated here.

[64]   This policy was expressly employed when formulating the information given to the Department. By internal email dated 5 September Mr Moran commented on Dassault’s answers to the long list of questions the Department had sent. In it he provided some “brief guidelines with respect to answering these customer questions as we are still operating in a sales cycle”. This included the advice to “refrain from making reference to customisation” in the answers, but to refer to configuration instead. He confirmed in cross-examination that the word customisation was to be limited to the base layer, and that they would never make changes to the base layer. He also confirmed this was an instruction to be followed by the technical people who were preparing the answers.

[65]   That approach had the capacity of being misleading, and was so misleading here. As Ms Gayner’s email exchanges show the Department’s requirements, even for stage one functionality, could not be met by the out of the box functionality and changes to the Quintiq product were required. Yet the representations were that the Quintiq product met all of the functionality requirements out of the box with no

customisation. I address further below the potential technical means of the expressions “customisation” and “configuration”.3 Whilst they have a technical meaning, I accept Dr Tan’s evidence that the differences between the expressions break down when used in particular contexts, and the difference between them is ultimately a matter of degree. I accept that the representations in the RFP, and surrounding the RFP, that no customisation was required for the Quintiq solution conveyed the meaning that no substantial changes to the standard functionality of the Quintiq product was required. That is so whether or not Dassault had an internal definition of the word “customisation” that it adopted to make it seem that its product could be provided without significant changes to its functionality. When a party uses technical language when dealing with another party it is what the language means to a reasonable recipient that matters.4

SOW23

[66]   The first contract between the parties relating to the Quintiq solution was entered on 3 September 2018. This was a contract between the Department and Fujitsu called SOW23. The pre-existing MSA between the Department and Fujitsu contemplated that particular statements of work (SOWs) could be entered into for other work, and SOW23 was one such statement of work. It was a contract for a stated price of $128,501 (plus GST). The price was on a time and attendance basis, but it was recorded that it was not expected to exceed the amount indicated. Two days later, on 5 September 2018 Fujitsu and Dassault entered DS SOW23, being an effective back-to-back contract between those parties.

[67]   The purpose of SOW23 was to obtain a better understanding and elaboration of the functional requirements of the Department, and also to undertake technical workshops to determine the feasibility of integrating SAP and Quintiq. Pursuant to the arrangements, workshops were held in the week of 10 September 2018, and further technical workshops were held in October 2018. The work in SOW23 revealed there

3 See [132]–[134] below.

4      See Gunton v Aviation Classics Ltd [2004] 3 NZLR 836 (HC) at [244]; West v Quayside Trustee Ltd (in Rec and Liq) [2012] NZCA 232, [2012] NZCCLR 16 at [30]; Anderson v De Marco [2020] NZHC 2979 at [84].

was some greater complexity with SAP integration. By the time of the Department’s TAC meeting on 25 October it was noted that:

Quintiq does not have an out of the box SAP integration. Due to this, some integration points will need to be built. Both SAP and Quintiq will provide the tools to complete the builds.

[68]   The position had been addressed at the integration workshops. The Department became aware through these processes that the integration with SAP would involve greater complexity and cost and that the full scope of the complexity would not become apparent until the analysis and design phase of the contract was undertaken. The Department also became aware that Dassault did not have SAP specific integration tools notwithstanding that the RFP response had stated that “Quintiq has a wealth of experience regarding the integration of Rostering Master Data and Time & Attendance Data to SAP payroll”. Given that SAP was a leading payroll software system and what the RFP stated, it would reasonably be expected the Quintiq solution would have a “out of the box” product to integrate with SAP. Dassault had a policy of not telling customers that it did not do so. In an email sent in January the following year one of its technical personnel, Mr Lee Ong referred to the fact that Dassault did not have an out of the box interface to SAP as “our internal fact”. Mr Ong confirmed in cross-examination that this was something they knew internally but that Dassault told customers something different. I accept that this was part of the misrepresentation of Quintiq’s out of the box functionality. Having said that, Dassault’s integration tools, whilst more generic, would allow SAP integration to occur. And in any event, the Department became aware that Dassault did not have specific SAP integration software at this stage.

[69]   In addition, and again as a consequence of greater knowledge gleaned through the work, the Department contemplated purchasing additional “modules” associated with the Quintiq product. In particular:

(a)An “equipment/asset” module. This was additional functionality that allowed equipment to be allocated to officers when they were allocated to tasks in the roster. Whilst this had been shown in the demonstrations it was not part of the first stage functionality that had been part of the RFP price for that stage.

(b)An “advanced leave management” module. This allowed the roster to be managed in light of the different types of leave entitlements held by officers.

[70]   The fact that additional “modules” were identified for purchase reiterated the representations, however. That is because they further confirmed that Quintiq was an “out of the box” solution. That is because the additional modules suggested that additional out of the box functionality could be obtained by purchasing the additional modules.

Revised pricing

[71]   In October 2018 Ms Stewart for the Department asked Mr Kathiresan of Fujitsu for updated pricing. She asked that this be provided by 16 October. She did so because the Department needed to know if there was any need to increase the overall MSW budget that would have to be approved at Ministerial level. This request was passed on by Fujitsu to Dassault.

[72]   As Ms Gayner’s earlier email exchanges show Dassault had planned to manage the expected increase in price. Mr Deans responded to the request for revised pricing in an internal email saying “we will have to justify the change in price” and asking Ms Ginevra Morgan to prepare a  list  of  reasons  why  the  price  had  increased.  Ms Morgan had not been involved in the original RFP response but she then provided “a high level list of the non-out of the box functional requirements” as the price increase justification. Ms Gayner then revised the list. She stated in her emails that two kinds of additional statements needed to be made to the Department. First, she said that assumptions about the functionality needed to be added to Dassault’s pricing and “this must be sent to Corrections to protect us down the track”. The second was the explanation Mr Deans had suggested to justify the price increase from the original RFP.

[73]   Ms Morgan’s initial list included items that were not departures from the original RFP. For example it included “organisation to working unit mapping” which is a reference to the Department’s use of working units in its operations. The fact that this could not be addressed by out of the box functionality is significant. But this

feature of the Department’s operations had been an issue that had been identified even before the RFP, including at the meeting on 16 February.

[74]   Ms Gayner’s revised list was then converted into a PowerPoint presentation. This stated, when describing the suggested changes to the proposed services that the Department was seeking:

The original RFP submission assumed an “out of the box” delivery approach with minimal configuration. After an initial kick off in recent workshops, it is now recognised that a significant amount of additional configuration has been requested in the following areas …

The list was then provided. This PowerPoint presentation was sent by Dassault to  Ms Horwood and Mr Wills of Fujitsu on 17 October 2018. Ms Horwood then converted what was in the PowerPoint presentation back into a word document. The “ballpark” revised figure associated with these documents involved the new price estimate of $1,825,811.20. This increase included the two additional modules that the Department had indicated it wanted to purchase.

[75]   The Department alleges it was not sent this increased price estimate or the reasons for it. There is no evidence that the PowerPoint presentation prepared by Dassault was sent by email to the Department. Neither has it been shown that any other document, such as Ms Horwood’s word document, was sent. Ms Stewart gave evidence for the Department that she was not provided with any revised pricing by Fujitsu other than an estimated increase in licence costs. Ms Horwood did not give any evidence that she provided the revised pricing to the Department. Mr Kathiresan, who had been copied into Ms Horwood’s email, confirmed that he did  not  do so.  Mr Wills was not called as a witness by Fujitsu. Late in the trial Fujitsu applied for leave to call Mr Wills as a witness. I declined the application for reasons set out in my minute of 27 September.5 Given Ms Stewart’s evidence, the lack of any documentary record of revised pricing being provided, and the lack of any evidence from Ms Horwood, Mr Kathiresan or any other witness from Fujitsu that revised pricing was provided I accept Ms Stewart’s evidence that the revised pricing estimate was not provided to the Department.

5      The Department of Corrections v Fujitsu HC Wellington CIV-2021-485-423, 27 September 2023.

[76]   There was, however, a separate PowerPoint presentation prepared by Dassault that Mr Wills sent to Ms Stewart dated 16 October in response to the request for revised pricing. This PowerPoint presentation was limited to outlining the two additional modules that the Department wished to purchase. It does not include revised pricing or explanations for increases in pricing overall. In addition by email dated 15 October Mr Wills sent Ms Stewart a list of revised prices for the Dassault licence, including a number of options. The fact that a document trail exists showing these exchanges further confirms that the revised pricing for stage one implementation that Dassault had prepared, and the explanations for it, was not passed on to the Department. In closing Fujitsu advanced a number of complicated arguments based on other contemporaneous documents to suggest that Ms Stewart’s evidence should not be accepted. I do not accept these arguments in the absence of more straightforward evidence or documentation.

[77]   I note that there is an internal Fujitsu email authored by Mr Wills which was put to Ms Stewart when she was recalled suggesting that an indication of an increased price was provided by Mr Wills orally. This document was not discovered until late in the trial, and I declined leave for Mr Wills to give evidence for the reasons addressed in my minute. Ms Stewart denied she had been provided with an oral update and I accept her evidence. The document alone does not show the revised pricing was so provided.6 In any event, even if some oral advice of a price estimate increase had been given, it would have been short of what was required. Fujitsu had an obligation to provide both the increased price, and the suggested reasons for it, to the Department. Fujitsu do not suggest that Dassault’s PowerPoint presentation with the increased price and the explanations for it were provided. Its only argument was that the price increase was passed on by Mr Wills to Ms Stewart over coffee — an argument I do not accept given the lack of any evidence to show this. The explanations that had been advanced for the increase in pricing were significant. Some of them could have been debated.

[78]   The evidence shows that Fujitsu had its own view on how the price increases could be managed with the Department, which explains why the pricing information

6      Taylor v Asteron Life Ltd [2020] NZCA 354, [2021] 2 NZLR 561 at [68].

was not provided at all at this time. So I accept that both Dassault and Fujitsu had strategies for dealing with the anticipated pricing increase.

[79]   The Department was nevertheless aware that the likely costs were increasing. Ms Horwood had indicated in August 2018 that the costs were increasing from the RFP response. Ms Stewart was doing her own work on the overall budget, and she used the information she was provided by Fujitsu in that process. Her own revised pricing was set out in spreadsheets that were created and amended in October 2018. In those spreadsheets the cost attributed to installing Quintiq was increased from the approximate figure of $700,000 from the RFP response to $1 million. Her figures also included the revised licensing costs.

[80]   What the Department’s own assessments demonstrate is that each of the Department, Fujitsu and Dassault knew that the project would cost more than estimated in the RFP response. If the claims that are made in this proceeding hinged on the difference between the Department’s estimates at this time compared with those that had been prepared by Dassault, a more detailed assessment of the differences between the assessments might be required. But, in any event, the real cost of Quintiq, as shown by the pricing estimates provided by Fujitsu/Dassault in 2019 are fundamentally higher than even Dassault’s revised pricing in October 2018. So the case does not depend on such a comparison.

[81]   Dassault’s PowerPoint presentation with the price estimate increase accurately described the RFP response as involving an out of the box solution with minimal configuration. It said, however, that this was an assumption notwithstanding it was not recorded as an assumption in the RFP response. It then recorded the suggested complexity that had developed leading to the higher price estimate, but even this was significantly lower than what proved to be the true cost of this project.

[82]   The more significant issue arising from the revised pricing estimates in October 2018 is the way in which price escalation was being managed by both Dassault and Fujitsu. Both of them were less than full and frank with the Department. Dassault’s presentation was part of an attempt to justify a price increase

notwithstanding what had been represented in the RFP, and Fujitsu did not pass on the price estimate increase or the justifications for it for similar reasons.

Gap between out of the box and required functionality

[83]   The key issue for Dassault was that there was a greater gap between its standard or out of the box functionality stated in the RFP response and what the Department had sought. A further emerging issue was that the Department had not yet contractually committed itself to the Quintiq solution. There had been a contract for preliminary work (SOW23) but the Department had not yet signed a contract for the implementation of the Quintiq solution, or purchased the licence for the Quintiq software. For Dassault the licence, in particular, was commercially significant. A contract for implementing the Quintiq solution involved charging on a time and materials basis for the work, but a contract for the licence involved no associated expenditure and represented a very significant element of Dassault’s potential profit from the project.

[84]   These two issues became related. The contract for implementation of the Quintiq solution was also a statement of work — SOW27. It was being drafted by the parties  during  September  and  October  2018.   By  email  dated  18 September   Ms Morgan suggested to the Dassault personnel, including Mr Deans and Ms Gayner that Dassault provide to the Department the Quintiq Product Description Document or “PDD”. This was a detailed document that described Quintiq’s product, including its out of the box functionality. It would potentially identify what was not in the standard offering, and accordingly what would need  to  be  configured/customised.  When Ms Morgan suggested it be provided to the Department she said providing it “would reduce risk and be very clear”. Ms Gayner responded in the following way:

Typically we would attach the PDD and then articulate the customer’s gaps in the SOW. We would then lock down the modelling effort etc on those gaps. Once you attach a PDD they will review it and be calling out everything that is a gap that is not included in the SOW.

My concern with Corrections is we do not know the gaps yet as we have not locked down all of the process and done a full review as things are still moving. Locking down anything too tight especially with government can significantly cause delay and put the project at risk. We did this at Qld Police and spent 6 months in an SOW review cycle for putting in too much detail. So given this SOW will only be for analysis (not modelling where the risk will

rise) and it’s a T&M engagement, my preference would be to not lock down the scope in too much detail in the SOW to ensure we get the license across the line which has already been extremely tough to achieve.

[85]   The references to getting the licence “across the line” and to putting “the project at risk” were references to the Department committing itself to purchasing the Quintiq licence. The issue was that identifying the gap between standard functionality and the Department’s requirements prior to the licence being purchased created a risk that the Department would not wish to proceed as the gap was too great and the cost of bridging it too high. In evidence Ms Gayner confirmed that the example she had referred to, Queensland Police, involved the customer withdrawing and not purchasing the licence when the position was investigated. Dassault wanted to avoid the Department also doing so.

[86]   This was a further aspect of the strategy to manage the concern arising from the RFP response. Given the representations that have been made in the RFP response, and reiterated in the demonstrations and the associated materials, I accept the Department’s argument that following this approach involved a conscious decision not to correct the misrepresentations by allowing the Department to identify the extent to which the Quintiq solution was not out of the box, and that it would require work of unknown scope to adapt it to meet the Department’s requirements.

[87]   I note that Dassault provided other technical documents concerning Quintiq during this period. They included what were described as “white papers” which were provided to the Department on 10 September. But I was not provided with evidence that explained the significance of these documents, or how they may have addressed the out of the box attributes of Quintiq.

[88]   During this period the Department also asked if Dassault would be prepared to allow the Department to enter a contract to implement the Quintiq solution without first purchasing the Dassault licence, or come to some other arrangement to defer the purchase of a licence cost, or part of that cost. From the Department’s perspective it was still to learn more about the Quintiq product and how it would be implemented, and it was expected that this would become clear in the analysis and design phase of

the project contemplated by the draft SOW27. Its preference was to defer committing to the substantial licence cost until such matters were clearer.

[89]   The original RFP response had stated that the Dassault licence would need to be purchased prior to such implementation work, however. There was some suggestion during the discussions that Dassault might be prepared to come to some arrangement, but this ultimately came to nothing. It was generally a key aspect of Dassault’s commercial operations to secure the licence payment as soon as possible. Moreover, in this particular situation, Dassault was faced with the problem of the apparent gap between what had been said in the RFP response and the true position. Exposing that gap before the licence was purchased could jeopardise the sale. In those circumstances Dassault made it clear that they were not prepared to depart from the requirement for the Department to purchase the licence before work began. Dassault also stated that its licence price would increase if a contract was not entered in December.

[90]   The Department’s lean RFP response had contemplated it having “off-ramps” where it could avoid contractual commitment as a consequence of what became apparent in the process, and there was an expectation all round that more detailed work would be done as part of contractual performance. But Dassault’s insistence on the Department committing to the licence cost meant that contractual commitment was required, and the Department was no longer able to rely on any “off-ramps”.

[91]   Dassault’s policy of avoiding providing information to the Department that would reveal the potential gap between its out of the box product and the Department’s needs was reflected in other ways. The draft contractual documents for what became SOW27 were being worked on by the parties at this time, and Dassault’s strategy of limiting the scope of the first release was reflected in the terms. In the back-to-back agreement between Fujitsu and Dassault — DS SOW27 — the clause identifying the scope of the services covered by the agreement stated:

[Dassault] assumes the Department will be following the Workforce planner product solution out of the box. Therefore, minimal configuration will be required.

[92]   Such an assumption was not recorded in SOW27 between Fujitsu and the Department, however. Ms Gayner was not able to explain why the agreements were different.

[93]   Neither do I accept that the draft SOW27 documentation otherwise corrected the misrepresentations that had earlier been made. It was formulated, particularly by Ms Gayner, in an attempt to reduce the scope of the stage one product the Department was obtaining, but it did not correct or withdraw the previous incorrect representations about Quintiq’s out of the box functionality. Emphasis was placed on one line included in the draft contract which stated that one of the assumptions was that “each Department correctional facilities sites will be mapped one-to-one as a working unit”. That assumption had earlier been recorded in a document recording the outcome of functional workshops in September. Fujitsu and Dassault argued that this added a qualification, or assumption, on the standard functionality of the Quintiq product would provide, and that accordingly that it did not deal with the more complex issues arising from units within the prisons. I do not accept this. It is not clear enough to operate as a significant qualification in this way. If this was an intended qualification of what had been a significant element of what was previously represented it needed to be more expressly raised. Moreover SOW27 was subject to the MSA, and accordingly it was subject to the representations in the RFP response.

[94]   It is apparent that the assumption was added at Mr Robertus Driessen’s suggestion given that it was foreseen that complications could arise. He was the senior technical architect at Dassault. But the fact that each prison site would be “mapped” in this way did not mean that Quintiq could not also address the units within those prisons with standard functionality. An important matter of this kind could not be addressed by such an oblique reference to a key feature that the solution was to address. It is of significance that Dr Tan, Dassault’s expert, gave evidence that he would have come close to “pulling the handbrake” on the project when it became apparent that the units issue was going to create complexity and greater cost.

[95]The relevant contracts were then signed in December 2018. This included:

(a)The Department and Fujitsu signing contracts varying the MSA to include the work described in the RFP response (involving Quintiq for the rostering solution for the Making Shifts Work programme) on 12 December 2018.

(b)Fujitsu and Dassault entering a “one-time reseller agreement” dated 12 December 2018 to enable Fujitsu to sell the Dassault licence, and the Department then acquiring the licence from Fujitsu.

(c)The Department signing a licence agreement with Dassault for the Quintiq  licence  and  associated  maintenance  and  support  dated   12 December 2018.

(d)The Department and Fujitsu entering SOW27 for the analysis and design phase on 18 December 2018.

(e)Fujitsu  and  Dassault  entering  the  back-to-back  DS  SOW27  on  20 December 2018.

(f)Fujitsu and Dassault entering a further sub-contractual agreement for the provision of application development services on 21 December 2018.

[96]   It is through these contractual documents that the Department contractually committed itself to the Quintiq solution by purchasing the Dassault licence, and entering the contract for the analysis and design of implementing it in the Department’s systems. Although the Department entered a licence agreement with Dassault it purchased the licence from Fujitsu at a cost of $1,596,531 (plus GST), and the contract in SOW27 involved an agreement for the Department to receive the analysis and design services on a time and attendance basis for an estimated total of $439,893 (plus GST).

Problems develop

[97]   When the parties sought to implement their contractual arrangements a number of problems emerged. Attempts were made by all parties to remedy the position. Fujitsu elevated the project status to red in its status report dated 5 March 2019. An escalation meeting was held on 24 April. No formal changes were made to the contract in accordance with the change request process in the MSA, however. Such changes were contemplated by SOW27. The Department also conducted reference checks on two other Dassault customers, Spotless and Falck in May and both reported complexities with integration issues. A further meeting between the parties was held on 8 May.

[98]   The extent of the problems ultimately led to the contractual arrangements being brought to an end on 21 June 2019 by the Department exercising its contractual right to remove the project as work under the MSA. There are a number of reasons why difficulties developed, but ultimately the key reason why the contract was brought to an end was the revised price estimates that Fujitsu provided on 28 May 2019.

[99]   I accept that there were other contributing factors, although for reasons I elaborate upon below I do not consider these matters to be significant to the claims that are made in this proceeding. In particular:

(a)Dassault’s approach, which Fujitsu was implementing, involved following what was called the Quintiq Project Life Cycle (QPLC). This was a methodology that Dassault applied to efficiently engage in the work required to implement Quintiq as required by the customer. For example it anticipated a customer representative be present at the meetings who would make the relevant decisions required for implementation. The expectations under the QPLC methodology was that the customer, here the Department, would follow the methodology. The QPLC process was referred to in the RFP response. I accept that the Department did not follow the QPLC methodology. I also conclude, however, that the methodology was destined to fail given the significant issues that had not yet been worked through by the parties, and the underlying significant gap between what the Department was expecting

and what Fujitsu intended to deliver for the estimated price that were to be confronted. The reality was that the parties were only just beginning the process of fully understanding both the Quintiq product and the Department’s requirements, and what was being revealed was a significant gap.

(b)The Department sent a large number of representatives to attend the relevant meetings. This involved a larger number than the QPLC methodology contemplated, and there was a lack of a single Departmental person then able to promptly make decisions at the meetings. The Department was sending a larger number of persons with operational knowledge in order to get a fuller understanding of Quintiq, to explain how the Department operated in practice, and in order to make decisions with workforce buy-in. Many of the Department representatives who attended had not undertaken the “learning modules” which were online exercises that supplied information about the Quintiq solution, how it worked, and how it would be implemented.

(c)Fujitsu’s project management, largely undertaken by Mr Kathiresan, was not as active or as solution focused as Dassault had expected. He acted more in an administrative capacity by making arrangements for the project, but as the substantial issues developed they became beyond his ability to manage. There was some criticism of him. There may be some validity in those criticisms — for example, I agree it is surprising that he had not undertaken the learning modules himself. But there were inherent problems with the project in any event. These inherent problems ultimately manifested themselves, and they were not due to a failure of project management.

(d)Work on SAP integration was delayed because the Department had failed to appoint an SAP integration contractor to address the issue from a SAP perspective. Integration with SAP required expert input by SAP specialists. During 2018 it was anticipated that this would be an

external party, but the Department was unable to find one. Ultimately it brought individuals in-house who had SAP experience to deal with SAP integration. This approach was less than ideal, and also involved delays. These were significant given the problems that SAP integration involved. In the end, however, whilst the Department can legitimately be criticised for this approach it was the underlying complexity with SAP integration that was a core problem rather than the delays in finding SAP expertise to assist in addressing them.

[100]   On 28 May 2019, following requests by the Department, Fujitsu provided revised pricing estimates. These were many times more than expected, or earlier represented. This involved the following prices:

(a)a “release one” price of $5,392,263 (with $1,652,651 being fixed price, and the remainder a “high level estimate”); and

(b)a “full scope” price of $7,158,149 (with $2,064,630 being fixed, and the remainder a “high level estimate”).

[101]   Ms Stewart gave evidence that many of the functional elements that were excluded from release one involved functionality that was included within the RFP proposal including:

(a)the requirement for work to be allocated to different cost centres;

(b)reporting functionality; and

(c)public holiday management.

[102]   Neither estimate compares precisely with the $716,000 price estimate in the RFP response. I consider the relevant comparison is significantly more than the

$5,392,263 price but not as high as the $7,158,149 price, which includes functionality over and above what was said in the RFP response. It is not possible to be more precise on the evidence available. But in any event the new estimated price was well beyond what had originally been estimated in the RFP response, and it was also significantly

greater than the Department had assessed before it entered the contracts in December 2018, or even Dassault’s revised pricing that it had provided to Fujitsu in October 2018 ($1,825,811.20).

[103]   It is to be noted that the May 2019 price estimates involved significant Fujitsu margins. The release one price estimate from Dassault was $3,338,761 and the full scope price was $4,544,807. But a Fujitsu margin was always contemplated, including in the original $716,000 price. In any event, the prices were well in excess of the previous estimates.

[104]   It was these revised prices estimates that ultimately led the Department to bring the project to an end by “de-scoping” it under the MSA by email dated 21 June 2019. Bringing such work to an end was contemplated by cl 8 of the MSA.

What went wrong?

[105]   Given the significant price estimates referred to above, which resulted in the project being brought to an end, it is clear that there was a major problem with using Quintiq as the Department’s rostering system. An important question is why that was so? That question is relevant to the claims of misrepresentation, false and misleading conduct, and breach of warranty.

[106]   Perhaps surprisingly the precise reasons why the Quintiq product did not work well with the Department’s systems and business practices, and required significant work to make it so operate, was not covered in much detail in evidence. Each of the parties called independent expert evidence. The experts had expertise in major technology projects of this kind. But none of them had expertise in the Quintiq product, or detailed knowledge of the Department’s systems. They accordingly could not give expert evidence on why the Quintiq solution did not work well for the Department.

(i)the extent to which a party was able to negotiate the terms of the agreement; and

(ii)whether a party was required to either accept or reject the agreement on the terms and conditions presented by the other party; and

(d)      whether the party seeking to rely on the effectiveness of a provision of the kind referred to in subsection (1) knew that a representation made in connection with the agreement would, but for that provision, have breached section 12A, 13, or 14(1); and

(e)      whether all or any of the parties received advice from, or were represented by, a lawyer, either at the time of the negotiations leading to the agreement or at any other relevant time.

…

Interpretation issues

[277]   There is no dispute that the pre-requisites in s 5D(3)(a)–(c)(i) are satisfied in this case. But there are interpretation issues concerning the provision that I need to address notwithstanding that they were not raised in submissions.

[278]   First the contractual clauses here include clauses that limit liability to a certain financial amount — particularly AUD10,000 in the Teaming Agreement and €100,000 (or the amount paid by Fujitsu in the 12 months prior) in the One Time Reseller Agreement. I consider such clauses to be of the kind referred to in s 5D(1). That is because they “indirectly” allow a party to engage in contravening conduct without an order being able to be made under s 43 beyond the stated amounts. They accordingly exclude the application of the Act beyond such amounts. The extent of the limit would then become one of the  circumstances of the agreement to  be considered under       s 5D(4). This may be important if the limit is low. Here the limits arising under both agreements are low. They effectively exclude any material application of the FTA.

[279]   Secondly, it seems to me that it is open for the Court to conclude that it was fair and reasonable to give effect to the clauses in excluding the FTA for some elements of a claim that is advanced, but not other elements. The section introduces a broad enquiry, and there is nothing that requires the Court to adopt an all or nothing approach. The Court is able to conclude that it is fair and reasonable for the clause to exclude the FTA in some, but not in all respects. So the circumstances of the

agreement may suggest that some elements of the claim should be protected by the exclusion clause, but there are other elements of the claim where it would not be fair and reasonable to so exclude operation of the FTA because of the circumstances.

[280]   Thirdly the reference in s 5D(4)(d) to a party knowing that a representation would, but for the exclusion provided for in the agreement, be a breach of the kind identified should be interpreted purposively. I do not consider that the party would need to know that particular provisions of the FTA would otherwise be breached, or even know about FTA liability at all. In my view it is sufficient that a party would know that it has engaged in conduct that makes it potentially legally liable. I consider the sub-section is focusing on whether misrepresentations, or other misleading conduct, were knowing ones.

[281]   I also agree with the view of Gault J in Williams v Tellen Systems NZ (2013) Ltd that the factors identified in s 5D(4) are mandatory relevant considerations.52 The listed considerations are not exclusive, however, as the Court is required to take into account “all the circumstances of the agreement” with the listed considerations being inclusive only.

Application of s 5D

[282]   I consider that the clauses in the Teaming Agreement apparently exclude the operation of the FTA beyond the limit in the manner contemplated by s 5D(1). I do not accept Fujitsu’s arguments that the clauses of the Teaming Agreement are inapplicable because they were only “backward looking”, and did not address future misrepresentations. Whilst that may be so of cl 7.8 — the entire agreement clause — cl 2.2 regulated the future provision of information, and cls 6.1 and 6.2 then addresses liability for misrepresentation under cl 2.2.

[283]   The general approach to be applied under s 5D is relatively well settled. Although it is in the context of the similarly worded provision in s 50 of the CCLA, the Supreme Court said in ANZ Bank New Zealand Ltd v Bushline Trustees Ltd that “… the task of the court is to assess whether in all the circumstances, it is fair and

52     Williams v Tellen Systems NZ (2013) Ltd [2021] NZHC 1199 at [117].

reasonable for an entire agreement clause to be conclusive between the parties.”53 Like the CCLA provision the overall purpose of s 5D is to allow parties to exclude liability in accordance with freedom of contract, whilst recognising that there will be some circumstances where it is unfair or unreasonable for a party to do so. The provision acts as a safety valve to ensure that the Act is not excluded when this would be unreasonable or unfair, even when parties are in trade. This may be particularly so when the other party was in a situation of greater vulnerability.54 The factors identified to in s 5D(4) are indicative of that. But unequal bargaining power is not a pre-requisite for establishing that it is fair and reasonable that the parties be bound by their provision.  It is just a mandatory consideration.  The only pre-requisites are those in  s 5D(3). The fairness and reasonableness of excluding the FTA will still depend on the overall circumstances. Nevertheless when there are arms-length commercial parties who are in an equally strong bargaining position the Courts have tended to give effect to the exclusion of the FTA.55

[284]   In terms of the factors in s 5D(4) there is nothing in the subject matter or value of the services under the Teaming Agreement that are of particular note by themselves. These were significant commercial contracts concerning the delivery of services entered in the ordinary course of business. The respective bargaining position of the parties was essentially equal, and neither party was required to accept or reject the agreement in any way that suggests unfairness. The parties were both legally advised. So all these factors are generally strongly in favour of Dassault’s argument that the agreement should be enforceable in its terms. Of the listed factors in s 5D(4) the only factor in Fujitsu’s favour is that set out in s 5D(4)(d). That is because Dassault became aware that its conduct was misleading, and it took steps to prevent the Department or Fujitsu becoming aware of this before the entry of the agreements in December 2018 to ensure that the contracts were not put at jeopardy.

53 ANZ Bank New Zealand Ltd v Bushline Trustees Ltd [2020] NZSC 71, [2021] 1 NZLR 145 at

[132] (footnotes excluded).

54 See PAE (New Zealand) Ltd v Brosnahan [2009] NZCA 611, (2009) 12 TCLR 626 at [15]; Brownlie v Shotover Mining Ltd CA181/87, 21 February 1992 at 31–32; Sipka Holdings Ltd v Merj Holdings Ltd [2015] NZHC 1980 at [56].

55 LSD 2017 Ltd v Landscaping Direct Ltd (In liq) [2022] NZCA 657; Williams v Tellen Systems NZ (2013) Ltd, above n 52.

[285]   Given these factors I accept Dassault’s arguments that it is fair and reasonable to give full effect to the exclusion of liability clauses in relation to Fujitsu’s claims arising out of its liability in relation to the amounts the Department paid to it (totalling

$637,548.06) and the amounts the Department paid to third parties (totalling

$1,402,323.50). Fujitsu must accept the commercial consequence of it deciding to be the prime contractor, and accordingly taking the commercial risk of this contract, including the risks reflected with the warranties it elected to give. Those risks included the liability of this kind, and there is nothing unfair or unreasonable in Fujitsu being held to the bargain that it struck with Dassault in this context. These were arms-length commercial parties, with equal bargaining strength, and there is no reason for their contract not to regulate such liability.

[286]   But notwithstanding that this is a very strong starting point in relation to all of the liability that Fujitsu faces, I have concluded that it is not fair and reasonable for Dassault to exclude the FTA in one respect — that is the liability arising from the licence fee for Quintiq which forms the Department’s claim against Fujitsu. That is because there are factors in relation to this potential liability, and Dassault’s conduct, that are of particular significance. In particular:

(a)The licence fee was paid to Dassault, and it represents a complete windfall. The licence cannot be used by the Department. It is the fee to licence software that has not been installed in the Department’s systems because the relevant installation contract has been brought to an end as a result of the misrepresentations. In effect the Department purchased a licence for nothing, and there has been a total failure of consideration in that sense. I note that in the licence agreement between the Department and Dassault the one situation where the Department could have recovered in a claim directly against Dassault is if the Quintiq product had failed to operate in the Department’s systems in accordance with its documentation, in which case the Department could terminate the licence and recover what it had paid for it.56 There is an

56     Dassault Systèmes Customer Licence and Online Services Agreement dated 12 December 2018, cl 6.1.

analogy between that situation, and the present situation as the licence is unusable.

(b)Unlike other payments made to Dassault there was no cost to it associated with the licence. Unlike the work that it did on this project which involved the time and attendance of its executives under SOW23 and 27 the licence fee represents pure profit. The claim is simply seeking that Dassault repay an amount paid to it for a licence that it is not using, and a service that it does not provide.

(c)The licence fee was Dassault’s intended gain from its misleading conduct. When Dassault appreciated that the RFP response had not accurately represented what could be provided for the price indicated, it embarked upon the strategy of not disclosing this. A primary motivation when doing so was to secure the licence payment — in  Ms Gayner’s email to Mr Deans and others on 18 September 2018 she suggested that Dassault not provide information that would identify the gaps “to ensure we get the licence across the line which has already been extremely tough to achieve”. Dassault knew that its conduct had been misleading — a factor identified in s 5D(4)(d) — and it decided not to be open about that with either the Department or Fujitsu so that it could secure this licence fee. It would be wrong as a matter of principle, and inconsistent with the policy of the FTA, to allow Dassault to succeed with this strategy.

(d)Fujitsu is largely an innocent party in relation to the liable misrepresentations. The untrue statements in the RFP response, and the negotiations, are attributable to Dassault. Fujitsu is responsible for them as the prime contractor, and because Dassault was its chosen technology sub-contractor. I do not accept Dassault’s argument that the Court should exercise its discretion to reduce any recovery because of Fujitsu’s own conduct. Any discretionary reduction for such matters is already addressed by the reduction arising from the operation of the

exclusion clauses.57 The only element of misleading conduct Fujitsu engaged in itself arises from its failure to pass on Dassault’s increased price estimate and the suggested reasons for it. But liability for the claimed losses arises irrespective of that feature of the case. Fujitsu is liable in this case because of Dassault’s conduct, and it is not fair and reasonable that Dassault can avoid liability altogether in the circumstances.

(e)The licence fee payment arose under separate contracts involving the One Time Reseller Agreement between Dassault and Fujitsu, and the Customer Licence Agreement between Dassault and the Department. These involve separate considerations to be addressed under s 5D(4). Indeed it may be that it is the One Time Reseller Agreement, and its exclusions, rather than the Teaming Agreement to which s 5D is applied in relation to this part of the claim. Fujitsu operated as a conduit for the licence fee under these agreements. It did not provide any services in connection with the licence, and it was simply operating as Dassault’s one-off reseller. The substantive seller of the licence, and the party receiving the benefit for the licence payment, was Dassault.

[287]   Fujitsu advanced an alternative argument that Dassault was directly liable to the Department under the FTA as a matter of accessory liability, and that there were no exclusion clauses between the Department and Dassault that would prevent that recovery. Whether the Court makes such orders under s 43 is a matter of discretion.58 I do not agree to exercising the discretion in this way. That approach would be inconsistent with the contractual framework agreed between the Department and Fujitsu, and between Fujitsu and Dassault. It is more appropriate to address the contractual framework chosen by commercial parties in trade in the manner contemplated by the FTA, especially s 5D. I do not consider it appropriate to use accessory liability to make orders under s 43 in a manner that circumvents the need to confront this statutory framework for assessing liability. But, in any event, approaching the case in that way would lead to a similar outcome in my view. It is

57     See Shabor Ltd v Graham [2021] NZCA 448, (2021) 16 TCLR 177.

58 At [57].

fair and reasonable that Dassault be liable for the amount of the licence fee paid to it, but not more. That is so whether applying the test under s 5D or the discretion under s 43.

[288]   Nevertheless for the above reasons, in relation to this particular aspect of the claim — represented by the Department’s successful claim for the licence fees in the amount of $1,836,010.65, I conclude that it is not fair and reasonable for the clauses to exclude FTA liability given these circumstances. I conclude that Fujitsu should succeed against Dassault for this amount.

[289]   That finding is subject to one complication. The amount of the Department’s claim may well be subject to a margin that Fujitsu included. I do not consider that Fujitsu can succeed against Dassault for the amount of its margin. The amount which represents Dassault’s liability to Fujitsu should be limited to the amount that Fujitsu paid to Dassault in relation to the licence the Department acquired. That amount has not been clearly identified, and I am unsure whether it is apparent from the evidence. I accordingly reserve leave to the parties to allow them to identify the relevant amount for which judgment should be entered for Fujitsu against Dassault given the above findings.

Destruction of documents

[290]   Fujitsu’s final claim against Dassault involves an allegation that Dassault improperly destroyed documents associated with the project that it was obliged to keep.

[291]   This allegation is based on the terms of yet another agreement entered between Fujitsu and Dassault dated 21 December 2018 under which Fujitsu appointed Dassault as its sub-contractor for the intended work on the project.59 Amongst the terms of the contract were the following obligations:

Information to be maintained by the Subcontractor

3.5The Subcontractor shall keep full, true and up to date records and documentation relating to the Services provided under this Agreement.

59     Sub-contract Agreement for the provision of application to development services.

3.6The Subcontractor will, upon request, provide the Principal with copies of or access to, proper, accurate, auditable and up-to-date records, technical information (collectively, “records”) relating to the Services. For avoidance of doubt, the Subcontractor will not be obliged to provide any records which would be deemed as commercially sensitive (e.g. human resources I staffing records with respect to Personnel providing the Services, commercial information relating to the actual costs of the Services).

[292]   This aspect of the claim has two relevant elements. First Fujitsu has argued that Dassault’s defence should be struck out, or that adverse inferences drawn because it has admittedly destroyed contemporaneous records that were relevant to the case in breach of r 8.3 of the High Court Rules 2016. Secondly Fujitsu alleges there has been a breach of this agreement as the seventh cause of action.

[293]   As to the first matter, I dealt with Fujitsu’s application to strike out Dassault’s defence prior to trial.60 I essentially repeat the findings in that judgment here. Dassault’s destruction of materials was no better explained at trial. It called one further witness, Mr Michael Tohu, the litigation support manager of its New Zealand solicitors. He had no knowledge of what had happened at Dassault, and no better explanation for Dassault’s document destruction, and why it was allowed to happen, was provided in evidence. But I have nevertheless not relied on drawing adverse inferences when making the findings above. I am satisfied that most, if not all relevant documentation has been retrieved in one way or another. Dassault’s discovery has included adverse documents. Some have been relied upon for the purpose of making findings. I am not satisfied that there is further information that has been unavailable to the Court because of Dassault’s document destruction policies and its failure to comply with the High Court Rules.

[294]As to the seventh cause of action I am not satisfied that a breach of cls 3.5 or

3.6 arises. I do not consider that materials such as email exchanges amount to documentation “relating to the services” under the agreement as contemplated by these clauses. The term “services” was defined to include services provided under specific statements of work. So the obligation does not relate to all documentation generally, but rather to documents relating to the particular services under statements of work. It relates to the work that Dassault actually did under SOW27 after the contracts were

60     Chief Executive of the Department of Corrections v Fujitsu New Zealand Ltd [2023] NZHC 1900.

put into place in December 2018, not internal email exchanges within Dassault in the period leading up to the agreement, or even necessarily internal emails occurring after the entry of the agreement. The documentation would need to relate in some way to the actual performance of SOW27. Fujitsu has not proved that any missing documents are of that character.

[295]   In any event there has been no loss proved to have arisen from any breach the agreement in this respect. Fujitsu appears to have recognised this in closing by suggesting the appropriate approach is to award Fujitsu indemnity costs. I am prepared to consider the relevance of this as part of the Court’s discretion as to costs, but I am not prepared to award indemnity costs as a remedy for breach of contract, and do not consider there are any principles that would result in this award.

[296]For these reasons this cause of action is dismissed.

Conclusion

[297]   The Department engaged in a major exercise to reform its system for rostering its officers over the prison network. In its response to the Department’s RFP Fujitsu represented that the Department’s rostering needs could be met by the standard, or “out of the box” functionality of a product offered by its sub-contractor, Dassault, called Quintiq. It stated that no customisation to this product would be required, although some configuration would be necessary in relation to particular matters. It represented that the Quintiq solution could provide the Department with what it needed for core functionality at an estimated price of $716,000 in addition to a licence fee.

[298]   The Department had only described its rostering needs at a very high level when these representations were made, but Fujitsu did not qualify these representations by suggesting that their accuracy depended on it having more information about the Department’s systems. The RFP response was subject to a very clear qualification, however, that it could not be relied on in any legally binding sense, and that all matters would need to be confirmed before contractual commitments were made. But in the period following the RFP response and the entry of the contracts in December 2018 the representations that Fujitsu and its sub-contractor Dassault had

made were reiterated, both in product demonstrations, and in other respects, and Dassault then insisted that contracts were entered. The only relevant qualification to the prior representations was that the Department became aware that the position about SAP integration was much more complicated than Fujitsu had said.

[299]   During this period of time Dassault also realised that what had been said in the RFP response was inaccurate, and that there would likely be a gap between the Department’s requirements and what could be provided with Quintiq’s standard functionality. But it decided not to correct the misrepresentation as it did not want to put the contract, and particularly the significant licence fee that the Department would commit to on entering the contract, at risk.

[300]   In any event, Fujitsu elected to provide warranties to the Department when it entered the contracts for the provision of the Quintiq solution in December 2018. Those warranties included a warranty that what had been said in the RFP response was accurate and had not changed in any materially adverse way. The qualification in the RFP response was accordingly no longer relevant.

[301]   What was stated by Fujitsu and Dassault about Quintiq was untrue. Quintiq’s standard functionality could not meet the Department’s needs, and extensive work in the nature of customisation was required for it to do so. The contract was brought to an end when Fujitsu provided an updated estimate in 2019 that the costs of doing so were in excess of $5 million — well beyond the $716,000 indicated in the RFP response for the core functionality that had induced the entry of the contract.

[302]   The Department succeeds in its claims for breach of warranty against Fujitsu, and is entitled to recover the amount it paid to Fujitsu for work to implement it, the amount it paid to Fujitsu for the Quintiq licence, and the third party costs it incurred in seeking to implement the contract.

[303]   Dassault is also potentially liable under the CCLA and FTA to Fujitsu because it engaged in misrepresentations and misleading and deceptive conduct. The misrepresentations in this case originated at Dassault, and it was Dassault that adopted the strategy of knowingly misleading the Department and Fujitsu. But the contractual

terms between Fujitsu and Dassault protect Dassault from liability. Parties are entitled to contract out of liability under the CCLA and FTA however, and Fujitsu and Dassault did so here. But that finding is subject to one qualification. Dassault cannot avoid its liability under the FTA to repay the licence fee that was charged for the Quintiq product. It would be grossly unfair, notwithstanding the exclusion clauses, to allow Dassault to keep this fee, and thereby allow it to secure a windfall gain, being the very gain it sought to achieve by knowingly misleading the Department and Fujitsu.

[304]   For the above reasons I uphold the Department’s claim against Fujitsu in the amount of $3,875,882.35 for breach of contractual warranty, and I uphold Fujitsu’s claim against Dassault under the FTA in relation to the amount of $1,836,010.65 less the amount of any margin charged by Fujitsu on that sum. I reserve leave to address the question of the margin, the appropriate calculation of the quantum, and in relation to interest. Those matters are to be addressed at the same time as costs if they cannot be agreed. Given the leave reserved this judgment is an interim judgment in accordance with r 11.2(a) of the High Court Rules.

[305]   I will receive memoranda of counsel in relation to costs if these cannot be agreed. It may be that there have been settlement offers that are relevant to costs. But by way of preliminary indication it seems to me that the starting point is that the Department is entitled to recover costs against Fujitsu, and Fujitsu is entitled to recover against Dassault. In relation to Fujitsu’s recovery against Dassault, however, it would only be in relation to its claim and not the full costs of trial. Whilst the Department will likely be able to recover costs for the whole trial it may be that only half the trial was attributable to Fujitsu’s claims against Dassault for the purpose of Fujitsu’s costs claim. If the position is not agreed I will see memoranda from the party or parties seeking costs within 30 working days of release of this judgment, and memoranda in response within 10 working days thereafter. All memoranda should be no longer than 12 pages plus a schedule. No memoranda in reply can be filed without leave (which may be sought informally by email to the Registrar).

Cooke J

Solicitors:

Bell Gully, Wellington for Plaintiff

Wigley and Company, Wellington for Defendant MinterEllisonRuddWatts, Auckland for Third Party