//
© 2026 Zoe. All rights reserved.
Zoe is a legal information platform. Always consult the Federal Register of Legislation for authoritative text.
[2025] NSWCA 178
respondent. Appeal dismissed. The appellant was ordered to pay the respondent’s costs. The primary judge’s declaration of beneficial interests in the proportions of two-thirds for the appellant and one-third...
The Court of Appeal upheld the finding that Scott Street was not beneficially owned solely by the person who paid its purchase price. The primary judge found that the parties deliberately bought it in joint names as part of a shared investment endeavour, and that Costaras had made substantial financial and non-financial contributions to their common wealth. When the relationship and endeavour ended, equity could account for those contributions and the parties’ circumstances. The two-thirds/one-third division was not shown to be erroneous. The Court also noted that AI-generated legal submissions can contain irrelevant material and invented cases, and stressed the need for careful verification.
AI-generated legal information, not legal advice. Zoe can make mistakes. Check the cited source, and for advice about your situation consult a qualified Australian lawyer.
6 cited instruments linked from this judgment.
2,272 words · generated 06/10/2026
May v Costaras [2025] NSWCA 178 concerned the beneficial ownership of an investment property in Scott Street, Maryborough, Queensland. Michael May and Lila Costaras were registered as joint tenants. The $180,000 purchase price was paid from May’s bank account, but the property was deliberately transferred into both parties’ names under the contract dated 30 June 2022 ([25]-[26]).
The property transaction took place within a wider relationship and property context. The parties began a relationship in late 2020 and lived first in Costaras’s rented home, then at May’s Cumberland Reach property. Costaras contributed to household expenses and later carried out substantial renovation-related work at Cumberland Reach. The Court’s account records that she had property-management experience, resigned from employment to manage and carry out renovation work, and contributed labour and project-management skills beyond ordinary household tasks ([22]-[23], [35], [42]).
In early 2022, the parties moved to Queensland and lived at Ann Street, a property May bought after consulting Costaras. The Scott Street property was then acquired as an investment. The primary judge found the parties expected renovation and sale, and that the Scott Street purchase was part of their wider shared economic arrangements. The relationship ended in September 2022 ([24]-[28], [37]).
May brought proceedings seeking to establish that Costaras held her registered interest in Scott Street on trust for him. Lindsay J found that the property was subject to a joint endeavour constructive trust and declared beneficial interests in the proportions of two-thirds for May and one-third for Costaras. May appealed, seeking 100 per cent or at least a 95 per cent share. The Court of Appeal dismissed the appeal with costs ([18]-[20], [45], [47], [93]).
The judgment also addresses Costaras’s use of Generative AI to prepare submissions. Bell CJ discussed the problems in a separate set of observations, while agreeing with Payne JA’s reasons and proposed orders ([1]-[17]). Those observations did not change the result of the property appeal.
The Court rejected the argument that May’s payment of the whole purchase price necessarily meant Costaras’s interest was held for him under a resulting trust. The primary judge had concluded that the facts, rather than a presumption, governed the dispute. In any event, the finding that the parties intended Costaras to be beneficially entitled from the time of purchase was inconsistent with the appellant’s claim that she was merely a nominee ([50]-[54], [59]).
A key part of the factual dispute was what the parties said before the contract was changed from May’s sole name to joint names. May argued that Costaras’s interest was conditional on their marrying. The primary judge did not accept that account, finding that the alleged condition lacked the certainty expected of a specific condition and that there was no concrete marriage plan or ceremony date ([55]-[56]). Costaras’s account was that her inclusion recognised her earlier contributions and the parties’ shared plans for another renovation investment. The primary judge preferred her account ([57]-[58]).
The Court of Appeal characterised the primary judge’s reasoning as a failed joint endeavour constructive trust, rather than a constructive trust based solely on common intention. Payne JA noted that the primary judge had discussed Muschinski v Dodds and Baumgartner v Baumgartner, and had addressed the existence of a joint endeavour, its breakdown without attributable blame, and what was required to do equity between the parties ([61]-[64]).
The Court accepted that the joint endeavour included Scott Street, which the parties acquired as a joint investment anticipating renovation and sale. The assessment of contributions was not confined to payments made directly towards Scott Street’s purchase price. The primary judge considered the parties’ common wealth, including Costaras’s work and expenditure at Cumberland Reach, her contributions associated with Ann Street, and the expected return from the Scott Street investment ([37], [41]-[44], [67]-[79]).
The appellant argued that the primary judge had failed to account for his own contributions to Cumberland Reach, had not identified each qualifying payment, and had not found that Costaras’s contributions generated a financial benefit to him. Payne JA rejected those complaints. The primary judge had expressly considered the parties’ respective contributions to their common wealth and had found Costaras’s renovation contribution to be in excess of $50,000 or thereabouts. The Court also explained, referring to Joudo v Joudo, that any common benefit need not necessarily be material in the sense of a financial benefit ([72]-[79]).
Finally, the Court rejected the challenge to the two-thirds/one-third division. The primary judge had considered both that May paid the purchase price and that Costaras was deliberately named as a co-owner in recognition of the parties’ shared endeavour and her contributions. The Court said the remedy was moulded to the justice of the case in an orthodox manner ([81]-[84]).
The judgment does not identify a change to the law of resulting trusts or joint endeavour constructive trusts. It applies existing equitable principles, principally the reasoning in Muschinski v Dodds and Baumgartner v Baumgartner. The point is how those principles applied to the particular findings about the parties’ relationship, contributions and intentions ([30]-[31], [62]).
Before the appeal, the primary judge had already found that the parties’ acquisition of Scott Street formed part of a failed joint endeavour and that it would be unconscionable for May to deny Costaras any beneficial entitlement. The primary judge also found it would be unconscionable for Costaras to retain an equal beneficial half share when May had paid the whole Scott Street purchase price and she had been relieved of further obligations to manage or renovate the property after being excluded ([43]-[44]).
After the appeal, those findings and the two-thirds/one-third division remained in place. The Court of Appeal did not replace the trial judge’s assessment with a new accounting exercise. It held that the appellant had not demonstrated legal or factual error in the approach taken ([81]-[84], [93]).
The judgment’s comments on Generative AI are distinct from its property-law reasoning. Bell CJ endorsed the need for appropriate oversight and verification, referring to observations in Ayinde v The London Borough of Haringey and to Practice Note SC Gen 23. The judgment does not establish a new substantive test for constructive trusts or a new formal rule governing AI use. It does, however, make clear that the Court viewed unreliable AI-assisted material as a practical concern for the administration of justice ([12]-[17]).
At [30], the judgment quotes the principle stated in Muschinski v Dodds: “equity will not permit the other party to assert or retain the benefit of the relevant property to the extent that it would be unconscionable for him so to do.” In plain English, where a shared endeavour fails, one party may not necessarily keep the benefit of the other’s contributions if doing so would be unconscionable. The quotation is part of the legal foundation for the joint endeavour analysis.
At [37], the primary judge’s finding is set out: Scott Street was acquired for a common purpose as a joint investment, with renovation and sale anticipated, and the parties intended the purpose to be pursued for their joint benefit. This matters because the case was not treated as a purchase where the source of funds alone decided beneficial ownership. The Court considered the shared purpose and the surrounding contributions.
At [54], Payne JA explains that the primary judge found an intention that both parties would be beneficially entitled “from the time of purchase.” This is central to the rejection of May’s resulting-trust argument. The finding was not merely that Costaras might acquire an interest in the future if the relationship continued or a renovation occurred.
At [72], the primary judge’s approach to remedy is quoted: the Court should give effect to practical considerations rather than pursue complicated factual inquiries that may produce insignificant differences, and should endeavour to do what is “practically just.” In plain English, the judgment accepts that equitable quantification may involve an overall assessment rather than a precise ledger of every payment.
At [86]-[88], the Court explains why it did not disturb the primary judge’s factual conclusion about the parties’ critical conversation. The trial judge had seen and heard both parties, and the text message relied on by May was consistent with Costaras’s account of having worked as part of a team to build a future. This illustrates the appellate court’s focus on demonstrated error, not simply a different possible reading of the evidence.
At [17], Bell CJ says the case illustrates the need for “judicial vigilance” in the use of Generative AI, especially, but not only, by unrepresented litigants. The immediate concern was that the respondent’s oral submissions included material that did not engage with the appeal and case references that were irrelevant or, in one instance discussed at [11], apparently non-existent.
The property-law reasoning is relevant to disputes where parties contribute money, labour or other resources to a shared relationship or project, acquire property in joint names, and later disagree about beneficial ownership after the shared endeavour ends. The judgment’s facts included a registered joint tenancy, one party’s payment of the entire purchase price, substantial contributions by the other party to earlier property renovation and household expenses, and findings that the later investment property was intended to benefit both parties ([25], [35]-[37], [42]).
The case is particularly instructive where a party argues that the purchase-money contribution should determine beneficial ownership through a resulting-trust presumption, but the evidence also includes deliberate joint registration and findings of a shared purpose. The Court’s conclusion depended on the primary judge’s findings about actual intention and the surrounding relationship, rather than treating payment of the price as conclusive ([50]-[60]).
The joint endeavour discussion may also be relevant where a claimed contribution is not a direct payment towards the disputed property. Here, the primary judge considered renovation work and expenditure relating to Cumberland Reach, and the Court upheld that wider assessment because the findings connected those contributions to the parties’ common wealth and the later Scott Street investment ([67]-[79]). The judgment does not say that contributions to any earlier property will always count. The connection to the particular joint endeavour and the evidence found by the trial judge mattered.
The appeal reasoning is relevant where a party asks an appellate court to re-evaluate credibility or contemporaneous messages. Payne JA emphasised that the trial judge had heard the witnesses and that no glaring improbability or other error had been shown ([85]-[89]). The treatment of falsified credit-card entries also shows that a trial judge may make an adverse credibility assessment on the evidence, even though the appeal court found no error in the approach taken here ([90]-[92]).
Separately, the AI observations are relevant to litigation conduct when a party uses a generative tool to prepare submissions or authorities. The judgment’s concrete examples involved irrelevant legal concepts, material that was not intelligible, and an authority list containing a citation to a case Bell CJ said did not exist ([4]-[11]). The observations concern risks of unverified material; they do not decide the property-law issues.
The supplied judgment does not discuss any later decisions applying or distinguishing May v Costaras. It was decided on 8 August 2025, and the source text provides no subsequent treatment. It would therefore be unsupported to describe a later line of authority or claim that another court has adopted, limited or rejected its reasoning.
The judgment itself discusses earlier authorities and explains their role. Muschinski v Dodds and Baumgartner v Baumgartner provide the equitable framework for the failed joint endeavour analysis ([30]-[31], [62]). Fox v Percy and Lee v Lee are cited in connection with the standard for appellate intervention in factual findings ([58], [85]-[86]). Joudo v Joudo is cited on the proposition that any relevant common benefit need not be financial in a material sense ([78]).
For the AI observations, Bell CJ discussed Ayinde and referred to Australian and overseas examples of erroneous AI-generated court material, including Olsen and Zzaman ([12]-[15]). Those are authorities discussed by the Court in relation to AI risks, not later decisions applying the property-law holding in May v Costaras.
The judgment does not establish a formula for calculating beneficial shares in all failed joint endeavour cases. It upholds the two-thirds/one-third division on the particular primary-judge findings and confirms that practical considerations may be relevant when moulding relief. The precise contribution assessment remains fact-dependent ([72], [81]-[84]).
The decision also does not state that every relationship involving joint property transactions is a joint endeavour, or that work on one property automatically gives rise to an interest in another. The result rested on findings connecting the parties’ contributions, common wealth and Scott Street investment purpose ([35]-[37], [67]-[79]). How a different evidential record would affect that connection is not determined here.
The judgment does not resolve a general question about when a resulting-trust presumption will apply in disputes involving joint registration and contributions made within a domestic relationship. It concludes that the primary judge was correct on the facts in finding no operative presumption or, alternatively, rebuttal ([53]-[60]).
On Generative AI, Bell CJ stresses verification and judicial vigilance, but the judgment does not set out a comprehensive procedure for litigants or courts. It endorses Practice Note SC Gen 23 and recognises that unverified AI use may add cost and complexity. The appropriate case-management response in other circumstances is not determined in this appeal ([13]-[17], [49]).
This is general legal information, not legal advice. Zoe can make mistakes. Before acting on this, relying on it, or using any draft, get a qualified Australian lawyer to check it.
Source: NSW Caselaw (www.caselaw.nsw.gov.au). Converted to plain text by ZoeLaw. Not the official version.
Medium Neutral Citation: May v Costaras [2025] NSWCA 178
Hearing dates: 30 July 2025
Date of orders: 8 August 2025
Decision date: 08 August 2025
Before: Bell CJ at [1];
Payne JA at [18];
McHugh JA at [95]
Decision: (1) Appeal dismissed.
(2) Appellant to pay the respondent’s costs.
Catchwords: EQUITY — Trusts — Resulting trust - whether primary judge wrongly failed to apply presumption that appellant did not intend respondent to obtain beneficial interest in property – whether respondent’s legal interest subject to purchase money resulting trust
EQUITY — Trusts — Constructive trust — joint endeavour constructive trust — de facto relationship — relationship broken down —investment property purchased in co-ownership as joint tenants — whether primary judge erred in characterising scope of the joint endeavour — whether the primary judge erred in determining beneficial entitlements of two thirds for appellant and one third for respondent — whether appellant entitled to larger beneficial interest
APPEALS — From finding of fact — Documentary evidence — text messages between the appellant and respondent — whether primary judge erred in making factual findings having regard to contemporaneous written records
APPEALS — From finding of fact — Credibility of witnesses — whether primary judge erred in assessing respondent’s credit
UNREPRESENTED LITIGANTS – Use of Generative AI – Hallucinated case reference – irrelevant case references – where Generative AI used to prepare script for oral submissions – where oral submissions so generated misconceived or irrelevant
Legislation Cited: Uniform Civil Procedural Rules 2005 (NSW)
Cases Cited: Ayinde v The London Borough of Haringey [2025] EWHC 1383 (Admin)
Baumgartner v Baumgartner (1987) 164 CLR 137; [1987] HCA 59
Burragubba v State of Queensland (2016) 151 ALD 471; [2016] FCA 984
Fox v Percy (2003) 214 CLR 118; [2003] HCA 22
Joudo v Joudo [2024] NSWCA 258
Lee v Lee (2019) 266 CLR 129; [2019] HCA 28
Mohareb v Saratoga Marine Pty Ltd [2020] NSWCA 235
Muschinski v Dodds (1985) 160 CLR 583; [1985] HCA 78
Olsen v Finansiel Stabilitet A/S [2025] EWHC 42Vernon v Bosley (No 2) [1999] QB 18
Zzaman v Commissioners for His Majesty’s Revenue and Customs [2025] UKFTT 00539
Texts Cited: Jacobs’ Law of Trusts in Australia (5th ed, Butterworths)
Category: Principal judgment
Parties: Michael May (Appellant)
Lily Costaras (Respondent)
Representation: Counsel:
B Phillips (Appellant)
Self-represented (Respondent)
Solicitors:
Proctor Phair Lawyers (Appellant)
Self-represented (Respondent)
File Number(s): 2025/129141
Publication restriction: None
[This headnote is not to be read as part of the judgment]
This matter concerns an investment property in Scott Street in Maryborough in Queensland (the “Scott Street property”), which is owned by the appellant, Michael May, and the respondent, Lila Costaras, as joint tenants.
The parties met in November 2020. They resided in the respondent’s rented Eastwood residence from about 10 January 2021, then moved into a property owned by the appellant at Cumberland Reach. In around February 2022, the appellant purchased a property located in Ann Street in Queensland. Shortly thereafter, the parties moved to Queensland and took up residence there, until the respondent moved out in September 2022.
In June 2022, the Scott Street property was purchased by the parties as joint tenants for $180,000. The payment of the purchase price of $180,000 was made from the appellant’s bank account. After the breakdown of the parties’ relationship, in 2022, the appellant commenced proceedings in the NSW Supreme Court seeking to establish that the respondent held her legal interest in the Scott Street property on trust solely for him.
The primary judge, Lindsay J, found that the parties held the Scott Street property on a joint endeavour constructive trust of the kind recognised in Muschinski v Dodds (1985) 160 CLR 583; [1985] HCA 78 and Baumgartner v Baumgartner (1987) 164 CLR 137; [1987] HCA 59. His Honour made orders declaring that the beneficial interest of the property was to be divided in the proportion of two thirds for the appellant and one third for the respondent. On appeal, the appellant claimed that he should have been awarded 100% or at least a 95% share of the Scott Street property.
The issues were:
(i) Did the primary judge err in not finding that the respondent’s legal interest in the property was held subject to a resulting trust in favour of the appellant?
(ii) Did the primary judge err in finding that that there was a constructive trust based upon a common intention?
(iii) Did the primary judge find a joint endeavour that only included the acquisition and development of the “Scott Street property” or one that included the acquisition and development of both the Scott Street property and the Cumberland Reach property?
(iv) Did the primary judge err in imposing a constructive trust based on a failed joint endeavour?
(v) Did the primary judge err in concluding that the parties held the Scott Street property in proportions of two-thirds and one-third?
(vi) Did the primary judge err by making the findings at [108]-[110] without having regard to contemporaneous written records relied upon by the appellant?
(vii) Did the primary judge err by failing to make findings about whether the respondent created false credit card records?
The Court per Payne JA, Bell CJ (with a separate concurrence) and McHugh JA agreeing held, in dismissing the appeal:
(i) The primary judge correctly found that the appellant had not established a basis upon which a presumption of resulting trust could operate, nor that, if such a presumption were applied, it had not been rebutted: [50]-[60].
Fox v Percy (2003) 214 CLR 118; [2003] HCA 22 and Lee v Lee (2019) 266 CLR 129; [2019] HCA 28 discussed.
(ii) The primary judge found that the facts established a constructive trust based on a failed joint endeavour, rather than a common intention. The appellant was correct to concede, in oral submissions, that the better characterisation of the reasons was that the primary judge found a failed joint endeavour constructive trust: [61]-[64].
(iii) It is clear that the acquisition and development of the Scott Street property formed part of the failed joint venture: [65]-[70].
(iv) The primary judge correctly imposed a constructive trust having considered that the parties acquired the Scott Street property for a common purpose (as a joint investment property, anticipating renovation and sale) and, allowing for past, present and prospective contributions to their common wealth, intending that their common purpose would be pursued, indefinitely, for their joint benefit. In so doing, his Honour sufficiently took into account contributions made by the appellant to the acquisition and development of the Cumberland Reach property; and the respondent’s substantial financial and non-financial contribution to renovation of the Cumberland Reach property: [71]-[80].
Muschinski v Dodds (1985) 160 CLR 583; [1985] HCA 78 and Baumgartner v Baumgartner (1987) 164 CLR 137; [1987] HCA 59; Joudo v Joudo [2024] NSWCA 258 discussed.
(v) The primary judge’s conclusion was informed by the respondent’s financial and non-financial contribution to the joint endeavour. Having explained the facts relevant to this case, the primary judge’s orders were moulded to the justice of the case in an orthodox manner: [81]-[84].
(vi) The primary judge concluded that the respondent was a “better historian” and that her version of the critical conversation was more likely to be correct. That finding has not been shown to be glaringly improbable. A close examination of the text messages relied upon by the appellant supports the conclusions reached by the primary judge: [85]-[89].
Fox v Percy (2003) 214 CLR 118; [2003] HCA 22 discussed.
(vii) The primary judge made a serious credit finding about the falsified credit card entries. No error was shown in his Honour’s findings: [90]-[92].
Per Bell CJ (Payne and McHugh JJA agreeing):
(viii) The respondent used Generative AI in the preparation of her oral submissions, many of which were not intelligible and did not engage with the matters raised on the appeal. This case illustrates the need for judicial vigilance in the use of Generative AI, especially but not only by unrepresented litigants: [1]-[17].
Ayinde v The London Borough of Haringey [2025] EWHC 1383 (Admin); Burragubba v State of Queensland (2016) 151 ALD 471; [2016] FCA 984; Mohareb v Saratoga Marine Pty Ltd [2020] NSWCA 235; Olsen v Finansiel Stabilitet A/S [2025] EWHC 42; Vernon v Bosley (No 2) [1999] QB 18; Zzaman v Commissioners for His Majesty’s Revenue and Customs [2025] UKFTT 00539 discussed.
BELL CJ: I have had the benefit of reading the reasons of Payne JA for dismissing the appeal. I agree with those reasons and the orders his Honour proposes.
I wish to add some brief observations to what his Honour has said at [49] below in relation to the respondent’s use of Generative AI in the preparation of her oral submissions in this Court. In so doing, I make no personal criticism of the respondent who was self-represented and doing her best to defend her interests. But the respondent’s use of Generative AI in the present case highlights the dilemma presented by the availability of that technology, especially where it is used by a person without legal training or otherwise not familiar with or unable to discern both the relevance and accuracy of what Generative AI may produce.
The respondent appeared remotely from Queensland by AVL and plainly was reading from a screen when delivering her oral submissions. The respondent was admirably candid with the Court in relation to her use of AI in the preparation of what she described as her speech for the purposes of the appeal hearing:
“BELL CJ: Ms Costaras, are you, and I don't mean this critically, I just want to inquire, are you reading from some script or some slides prepared by artificial intelligence?
RESPONDENT: Yes, your Honour, I did get the help of AI.”
The respondent later indicated that she had “uploaded all the documentation”.
Some of the respondent’s oral submissions were intelligible and engaged with the matters raised on the appeal. Many of them, however, did not. It is useful to give some examples to highlight the serious shortcomings of the use of Generative AI at least by a person who is not capable of either checking the accuracy or veracity or relevance of what has been generated. Again, I emphasise that in setting out the following passages, I am not being personally critical of the respondent who was doing her best to represent herself.
The following is an extract from the respondent’s oral submissions as recorded in the appeal transcript:
“The Scott Street property was intended to continue this model. The respondent sourced it, undertook arrangements for trades and quotes and immediately began preparations for the renovations before being locked out by the appellant he went to rescind counter restitution of the earlier joint endeavours. The defence also destroys a plaintiff's claim to an equitable proprietary interest in property, to which the defendant has legal title, Nisbet and Potts' Contract "The rationale of the application of good faith purchase in equity is to determine priority of title between competing claimants to property." Wheatley v Bell.
Moreover, as Lord Goff explained in Lipkin v Gorman, sorry, in Lipkin Gorman v Karpnale, the requirements which must be proved in order to establish the former also differ because change of position will only avail a defendant to the extent that his position has been changed. Whereas if a bona fide purchase is involved, no inquiry is made in most cases into the adequacy of the consideration.”
I make the following observations. First, the expression “he went to rescind counter restitution of the earlier joint endeavours” does not, on its face, make sense although what was intended may be capable of being inferred in the context of the parties’ dispute. But the introduction of notions of counter-restitution and restitution and the reference to Lipkin Gorman v Karpnale [1991] 2 AC 548 and change of position had absolutely nothing to do with the legal issues in the case.
Next, the reference to Nisbet and Potts' Contract, presumably a reference to In re Nisbet and Potts' Contract [1905] 1 Ch 391, was wholly inapposite to any of the factual or legal issues in the case as the headnote to the report of that case makes apparent. In that headnote, Farwell J’s decision is summarised as follows:
“Held, that a title acquired by adverse possession is not paramount to, and does not destroy, the equitable right of persons entitled to the benefit of prior restrictive covenants to enforce them against the land; that any purchaser taking less than a forty years' title is fixed with constructive notice of everything of which he would have received actual notice if he had insisted on a full title; and, therefore, the vendor in this case had notice of the covenants; and that a good title was not shewn.”
Next, the reference in the submissions to Wheatley v Bell was, it may be assumed, a reference to Helsham CJ in Eq’s decision of that name reported at [1982] 2 NSWLR 544. That was a case involving an injunction to restrain a breach of confidence in which his Honour held, inter alia, that there are no property rights associated with an equity to restrain a person from acting in breach of confidence which is owed to another, so that the equitable defence of bona fide purchaser for value which is directed towards the resolution of priorities in relation to property rights, does not apply: at 549-550. That decision had nothing remotely to do with the issues in the present case.
Another example of the misconceived, unhelpful and irrelevant oral submissions evidently produced by Generative AI was the following:
“The joint tenancy registration reflects an intention for survivorship, consistent with the relationship involving trust, reliance and joint economic life. It has been argued that a resulting trust would be imposed upon proof of one of the established grounds of restitution, provided that at the time of the claim the recipient has title to the plaintiff's property or its traceable proceeds, and that ground of restitution was not established before the recipient acquired full beneficial title to the property.”
“I appreciate you are representing yourself et cetera, but to the extent you're just relying on something which has been produced by AI, from what you've read out, it is not really engaging with the arguments. That doesn't mean that the Court is against you by any means. But there's limited utility in just reading out something which has been produced—"
It should also be noted that the respondent’s list of authorities identified under the heading “Cases to be read” the case of Tate v Ragg [2004] NSWCA 306: [66]-[70]. No such case exists. On the assumption that the respondent drew on Generative AI to produce her list of authorities, it is a matter of profound concern that the technology relied upon generated a non-existent case including pinpoint paragraph references to apparently relevant paragraphs. This has been and remains a serious issue: how does Generative AI produce facially credible citations to non-existent cases, still less provide paragraph references to such cases? And if it does so, what reliance can be placed on other legal references or propositions so produced? In addition, as Payne JA has pointed out, while the citation references to many other cases on the respondent’s list of authorities were not inaccurate, most had little, if anything, to do with the issues in this case.
After acknowledging that “[a]rtificial intelligence is likely to have a continuing and important role in the conduct of litigation in the future”, Dame Victoria Sharp, President of the King’s Bench Division of the High Court of Justice has recently observed, in delivering the reasons of the Court in Ayinde v The London Borough of Haringey [2025] EWHC 1383 (Admin) at [5]-[9] (Ayind****e) (omitting footnotes):
“This comes with an important proviso however. Artificial intelligence is a tool that carries with it risks as well as opportunities. Its use must take place therefore with an appropriate degree of oversight, and within a regulatory framework that ensures compliance with well-established professional and ethical standards if public confidence in the administration of justice is to be maintained. As Dias J said when referring the case of Al-Haroun to this court, the administration of justice depends upon the court being able to rely without question on the integrity of those who appear before it and on their professionalism in only making submissions which can properly be supported.
In the context of legal research, the risks of using artificial intelligence are now well known. Freely available generative artificial intelligence tools, trained on a large language model such as ChatGPT are not capable of conducting reliable legal research. Such tools can produce apparently coherent and plausible responses to prompts, but those coherent and plausible responses may turn out to be entirely incorrect. The responses may make confident assertions that are simply untrue. They may cite sources that do not exist. They may purport to quote passages from a genuine source that do not appear in that source.
Those who use artificial intelligence to conduct legal research notwithstanding these risks have a professional duty therefore to check the accuracy of such research by reference to authoritative sources, before using it in the course of their professional work (to advise clients or before a court, for example). Authoritative sources include the Government’s database of legislation, the National Archives database of court judgments, the official Law Reports published by the Incorporated Council of Law Reporting for England and Wales and the databases of reputable legal publishers.
This duty rests on lawyers who use artificial intelligence to conduct research themselves or rely on the work of others who have done so. This is no different from the responsibility of a lawyer who relies on the work of a trainee solicitor or a pupil barrister for example, or on information obtained from an internet search.
We would go further however. There are serious implications for the administration of justice and public confidence in the justice system if artificial intelligence is misused. In those circumstances, practical and effective measures must now be taken by those within the legal profession with individual leadership responsibilities (such as heads of chambers and managing partners) and by those with the responsibility for regulating the provision of legal services. Those measures must ensure that every individual currently providing legal services within this jurisdiction (whenever and wherever they were qualified to do so) understands and complies with their professional and ethical obligations and their duties to the court if using artificial intelligence. For the future, in Hamid hearings such as these, the profession can expect the court to inquire whether those leadership responsibilities have been fulfilled.”
I would endorse these observations. The great care that is required by, and responsibility of, legal practitioners in New South Wales in the use of Generative AI is reflected in Practice Note SC Gen 23 – Use of Generative Artificial Intelligence (Gen AI).
The appendix to the judgment in Ayinde contains non-exhaustive examples from England and Wales, the United States, Australia, New Zealand and Canada of material being placed before courts that has been generated by an artificial intelligence tool, but which is erroneous. At least two of those cases involved unrepresented litigants: Olsen v Finansiel Stabilitet A/S [2025] EWHC 42 (KB); Zzaman v Commissioners for His Majesty’s Revenue and Customs [2025] UKFTT 00539 (TC). The list of such cases continues to grow.
The problems of unverified use of artificial intelligence in the preparation of submissions are exacerbated where the technology is used by unrepresented litigants who are not subject to the professional and ethical responsibilities of legal practitioners and who, while subject to the Practice Note SC Gen 23, may be unaware of its terms. All litigants are under a duty not to mislead the court or their opponent: Vernon v Bosley (No 2) [1999] QB 18 at 37, 63, cited in Burragubba v State of Queensland (2016) 151 ALD 471; [2016] FCA 984 at [228]; see also, in relation to the obligations of unrepresented litigants, Barton v Wright Hassall LLP [2018] UKSC 12; [2018] 1 WLR 1119 at [18] (Barton). As Lord Sumption observed in Barton at [18], in a passage quoted in Mohareb v Saratoga Marine Pty Ltd [2020] NSWCA 235 at [39], “[u]nless the rules and practice directions are particularly inaccessible or obscure, it is reasonable to expect a litigant in person to familiarise himself with the rules which apply to any step he is about to take.”
It is and will remain important for judicial officers to be conscious of the potential use of Generative AI by unrepresented litigants in legal proceedings and it is legitimate to inquire, as the Court did of the respondent in the present case, whether Generative AI has been used in the preparation of materials placed before the Court. Such use may introduce added cost and complexity to the proceedings and, where unverified, add to the burden of other parties and the Court in responding to it.
At least at this stage in the development of the technology, notwithstanding the fact that Generative AI may contribute to improved access to justice which is itself an obviously laudable goal, the present case illustrates the need for judicial vigilance in its use, especially but not only by unrepresented litigants. It also illustrates the absolute necessity for practitioners who do make use of Generative AI in the preparation of submissions – something currently permitted under the Practice Note – to verify that all references to legal and academic authority, case law and legislation are only to such material that exists, and that the references are accurate, and relevant to the proceedings.
Although there was a challenge (Uniform Civil Procedural Rules 2005 (NSW) r 51.36 statement) made to the acceptance of a critical conversation by the primary judge, most of the facts found by the primary judge were not challenged.
The appellant and respondent first met on 22 November 2020, at a social event, described by the appellant as a “Facebook Page Single Parents Day event”. They commenced a romantic relationship on or about 23 December 2020. The appellant moved into the respondent’s rented Eastwood residence on or about 10 January 2021 and soon thereafter into a property owned by the appellant at Cumberland Reach. They became engaged to be married on or about 30 January 2021 when the appellant proposed to the respondent in front of her family and friends. There was no evidence that a date was ever fixed for a marriage ceremony to take place. The primary judge accepted that the parties shared a commitment to life as a couple. At the time they began their relationship, the appellant was 47 years old and had four adult daughters, no longer living with him. The respondent was 41 years old and had two adult children and two young daughters of primary school age.
In February 2021, the respondent moved into the Cumberland Reach property. For about four months after that time the respondent contributed to household expenses by fortnightly payments of $800. That the parties’ arrangement was essentially domestic rather than purely commercial was confirmed by the appellant’s evidence that at about the time the respondent made her last fortnightly payment he said to her: “I do not want you paying me rent. If you want to contribute you can pay for groceries and, if you can, the energy bill”.
In or about February 2022, the appellant (in consultation with the respondent who found the property) purchased a property located in Ann Street, Maryborough in Queensland (the “Ann Street property”). Shortly thereafter, the parties moved to Queensland and took up residence there. The appellant purchased the Ann Street property by drawing down a loan secured against the Cumberland Reach property.
By contract dated 30 June 2022, the Scott Street property, which is at the heart of these proceedings, was purchased by the parties as joint tenants for $180,000. The payment of the purchase price of $180,000 was made from the appellant’s bank account. The primary judge found that whilst there was some confusion about the appellant's banking arrangements which funded the purchase, the Scott Street property was purchased without any mortgage debt charged against it. The property has been rented out since its acquisition. As I will explain, the Scott Street property has now been sold and the primary judge made orders for the disposition of the net proceeds of sale. Settlement is to occur on 11 August 2025.
The contract dated 30 June 2022 was preceded by a contract dated 4 June 2022 (the “first contract”) in which the appellant alone was named as the buyer of Scott Street. That first contract was rescinded by an undated document styled “Deed of Rescission” made between the sellers and the appellant which provided for the property to be purchased by the appellant and the respondent (as “buyers”) in a “second contract”, entry into which effected a recission of the first contract pursuant to the Deed.
[59] In these proceedings, the significance or otherwise of the parties’ acquisition of the Scott Street property as joint tenants has not been explored. Nor is there any evidence before the Court of steps taken, or intended to be taken, to sever the joint tenancy. Nevertheless, the fact that the Scott Street property was purchased in the names of both parties as joint tenants is consistent with the currency of a relationship in the nature of a marriage.
At the heart of the primary judge’s reasons was his conclusion that the respondent “was not named by the plaintiff as a co-owner [of Scott Street] by accident or by oversight, but deliberately and mindful of her expectation of co-ownership arising from her proven contribution to renovation of the Cumberland Reach property and a common expectation that the Scott Street property would be the subject of a like restoration project” (J[9]).
As the primary judge explained, the “joint endeavour constructive trust” is one species of trust. In Baumgartner v Baumgartner, the majority (Mason CJ, Wilson and Deane JJ) at 148 adopted the analysis which Deane J (with whom Mason J had agreed) had advanced in Muschinski v Dodds (at 17):
… The principle operates in a case where the substratum of a joint relationship or endeavour is removed without attributable blame and where the benefit of money or other property contributed by one party on the basis and for the purposes of the relationship, or endeavour would otherwise be enjoyed by the other party in circumstances in which it was not specifically intended or specially provided that that other party should so enjoy it. The content of the principle is that, in such a case, equity will not permit the other party to assert or retain the benefit of the relevant property to the extent that it would be unconscionable for him so to do: cf. Atwood v Maude (1868) LR 3 Ch application 369 at 374-375 and per Jessel MR, Lyon v Tweddell (1881) 17 Ch D 529 at 531.
… by applying the general equitable principle which restores to a party contributions which he or she has made to a joint endeavour which fails when the contributions have been made in circumstances in which it was not intended that the other party should enjoy them.
In the present case, the only witnesses who gave evidence at the hearing before the primary judge were the appellant and the respondent. The primary judge summarised the parties’ respective cases at trial as being:
The appellant “says he agreed to purchase the property in joint names only upon a condition (to which the [respondent] agreed) that the parties marry” (J[20]).
The respondent says that “they agreed that the property be purchased in their joint names in recognition of their existing relationship, the contribution she made in renovation of the Cumberland Reach property, and the purchase of the property as a new joint renovation project. She says that she told the [appellant] that she would not remain in their relationship if not acknowledged as a co-owner of the Scott Street property. On her version of events her inclusion on the title to the property was not conditional upon marriage” (J[21]).
The principal issue litigated by the parties was the terms of a conversation which took place in mid-June 2020, prior to the purchase of the Scott Street property as joint tenants. The primary judge preferred the respondent’s account of events and observed that she was “a better historian” (J[101]). His Honour found that the appellant’s evidence of a conversation with the respondent in which the joint purchase of the Scott Street property was expressly conditioned upon their subsequent marriage was a rationalisation of events rather than a reliable account of what actually was said.
The primary judge did not focus upon whether the relationship between the parties was formally to be characterised as a “de facto relationship”, but made findings about the intention of the parties, and in particular the intention of the appellant, in purchasing the Scott Street property as joint tenants. The primary judge found that the respondent’s insistence upon, and expectation of, being named as a co-owner of the Scott Street property was, in all the circumstances of the case, not unreasonable.
This was because the respondent’s contribution to the renovation of the Cumberland Reach property involved extensive physical labour and project management work far beyond something nominal. The respondent had experience and skills in, and an aptitude for, property management that the appellant did not. The respondent gave up full-time employment in order to provide her project management services to what the primary judge found was a joint endeavour. The respondent’s contribution was consistent with a relationship such as that between an established married couple, de facto or de jure. The primary judge described the nature of the relationship between the appellant and the respondent thus:
[12] However one views the relationship it was not simply a “romantic” relationship, although that it was. Both parties wanted a stable, productive home environment and looked to property renovation projects as a means of mutual profit. The defendant had experience and skills in, and an aptitude for, property management that the plaintiff did not. He had worked hard as a truck driver and was looking forward to semi-retirement. From different perspectives they both hoped, and expected, to profit from renovation projects conducted as a joint endeavour founded upon a domestic arrangement in the nature of a marriage and in anticipation of a formal marriage at some indefinite future time.
[106] … on the whole of the evidence, the plaintiff made a deliberate decision (not conditional on a future marriage but in the context of an existing de facto relationship in the nature of marriage) to purchase the Scott Street property in the joint names of the parties with the intent that, from the time of purchase, they both be beneficially entitled to the property as co-owners, in recognition of their then current relationship and wanting it to continue, in acknowledgement of the promises he made to the defendant about renovation of Cumberland Reach property and the prospects of other renovation projects, in recognition of an agreement for the acquisition of the Scott Street property as a renovation investment for both parties, and with a shared expectation that the defendant’s acquisition of a beneficial interest in the Scott Street property would compensate her for her past and future renovation work. The parties purchased the Scott Street property with the actual common intention that from the outset the defendant would be a co-owner, not merely a nominee of the plaintiff, in deployment of the Scott Street property as a shared investment.
[117] On the facts of this case, the parties acquired the Scott Street property for a common purpose (as a joint investment property, anticipating renovation and sale) and, allowing for past, present and prospective contributions to their common wealth, intending that their common purpose would be pursued, indefinitely, for their joint benefit. Shortly after acquisition of the property their “joint endeavour” ended prematurely when they each came to realise that their personal relationship was not sustainable.
[118] The premature termination of their joint endeavour left each party with a disputed claim to beneficial entitlement to the Scott Street property unable to be resolved without the intervention of equity.
The primary judge noted that “[t]he larger question is how to resolve the parties’ competing claims to beneficial entitlement to the property in light of the irretrievable breakdown of their relationship and, consequently, their joint endeavour” (J[84]).
Before the primary judge, the appellant claimed a declaration that the defendant’s interest in the Scott Street property is held “subject to a resulting and/or constructive trust in favour of the plaintiff” and, in the alternative, a declaration that the legal interest in the property is subject to “a constructive trust to repay the contribution by the plaintiff to the acquisition and maintenance of the property (plus interest on those amounts), and as to the residue of such shares as the Court may deem fit”.
As regards constructive trust, the primary judge posed the starting point of analysis of the case as being that the respondent, without any allegation of fraud, was a registered proprietor of the Scott Street property. The ultimate question for determination was whether (upon an irretrievable breakdown of their relationship) it would be unconscionable for her to insist upon her legal entitlement to the property as a co-owner in circumstances in which the plaintiff paid the whole of the purchase price for the property.
In determining the beneficial entitlements of both parties, the primary judge found it necessary and appropriate:
to take into account the financial and non-financial contributions made by the respondent to the parties’ common wealth (largely by the expenditure of money and labour on renovation of the Cumberland Reach property), together with an allowance for her expected return on the joint investment in the Scott Street property, from which she was excluded by the appellant (J[120]); and
to give effect to practical considerations, rather than to pursue complicated factual inquiries “which may result in relatively insignificant differences in contributions and consequential beneficial interest” (Baumgartner at 150) and, in the formulation of an equitable remedy, to endeavour to do what is “practically just” (J[122]).
The primary judge emphasised that the respondent’s contribution to the joint endeavour and the “material welfare of both parties” was substantial, both financially and non-financially. She contributed to the expense of renovation works to the Cumberland Reach property and expenses associated with the Ann Street property and to household expenses while the parties were living at the Cumberland Reach and Ann Street properties. She also resigned from her employment as a local government Building Services Officer to manage and carry out renovation works to the Cumberland Reach property. It is not in dispute that her non-financial contribution went well beyond household chores. She personally performed work that required skill and care in management of renovation work and in location of the Ann Street and Scott Street properties. Whatever value might be attributed to her financial contributions, it was more likely than not that she devoted to renovation of the Cumberland Reach property an amount of funds not only substantial in itself, but a substantial proportion of the funds available to her. As such, the primary judge found that:
[107] … the plaintiff’s contention that it would be unconscionable for the defendant to maintain an interest in the Scott Street property fails to take into account, on a legal analysis, that the defendant took up with him, twice moved, quitted employment and relied upon promises of material benefit from renovation projects only to have him unilaterally terminate their relationship (both personal and economic), denying her entitlement to any part of promised material benefits and opportunities for profit, and excluding her from any enjoyment of the Scott Street property (of which she was, and remains, at law a co-owner).
On the other hand, the primary judge noted that in taking possession of the Scott Street property upon completion of its purchase and excluding the respondent from the property, the appellant relieved her of any obligation to contribute further to renovation or maintenance of the property pending a determination of their respective beneficial entitlements.
The primary judge thus concluded at [124] that:
(a) it would be unconscionable for the plaintiff to deny the defendant a beneficial entitlement in the Scott Street property (notwithstanding his funding of the purchase price), purchased as the property was as a vehicle for recognising the defendant's past, present and prospective contribution to the parties' common wealth in circumstances in which he has excluded her from the enjoyment of the property to which she is entitled as a joint tenant; and
(b) it would be unconscionable for the defendant to insist upon retaining beneficial ownership of a full one half share of the Scott Street property in circumstances in which the whole of the purchase price for the property was paid by the plaintiff and (rightly or wrongly) she has been relieved of any obligation to assist in management or renovation of the property.
- NOTE the reasons for judgment published to the parties as May v Costaras [2025] NSWSC 90.
- DECLARE that the plaintiff and the defendant each hold their legal interest in the land at 9 Scott St, Maryborough in the State of Queensland and described as Lot 2 in Registered Plan 3893 (Property) upon trust and in the proportions of two-thirds for the plaintiff and one-third for the defendant.
3. The plaintiff and defendant must each forthwith do all things necessary to cause the prompt sale of the Property, such things to include (without limitation:
(a) list the Property for sale with an agent agreed upon by the parties and failing agreement within 7 days of the date of this order, (a) the defendant must within 7 days thereafter provide to the plaintiff in writing, a list of three real estate agents; and (b) the plaintiff must within a further seven days select one from the list and notify the defendant accordingly;
(b) the parties are to provide to the agent a written authority and direction for the agent to send directly to both parties all information, communications and copies of documents relevant to the sale of the Property;
(c) the parties are to execute all documents requested by the agent required for the sale of the Property within 24 hours of the request being made, and if there is any dispute about the terms of the agent's contract within seven days of the agent being selected, the contact is to be in the agent's standard terms and with the agent's standard fees;
(d) give such instructions as are necessary to conveyancing solicitor agreed upon by parties within seven days of the Agent being selected and failing agreement, the plaintiff must forthwith in writing nominate three solicitors from which the defendant must within a further seven days select one and notify the plaintiff (Solicitor);
(e) the parties are to provide to the Solicitor an authority and direction for the Solicitor to send directly to both parties all information, communications and copies of documents relevant to the sale of the Property;
(f) instruct the agent to market the Property for sale by public auction on a date within ten weeks of the date of the selection of the agent (First Auction) at a reserve price agreed between the parties and failing agreement at the reserve price suggested by the agent;
(g) if the Property does not sell at the First Auction, market the property for sale with the Agent by way of private treaty for a period of 12 weeks during which time the parties must accept any offer made to purchase the Property within 5% of the reserve price of the First Auction unless the plaintiff and defendant otherwise agree in writing;
(h) if the Property is not sold at the First Auction and is not sold in the period provided for sale by private treaty, market the Property for sale by public auction with the Agent on a date within six weeks of the date of the conclusion of the period of sale by private treaty at a reserve price agreed between the plaintiff and defendant and failing agreement at 5% below the reserve price at the First Auction;
(i) in the event that the reserve price set for an auction is not reached, the defendant is to negotiate with the highest bidder and the second highest bidder and accept the highest offer to purchase made within 5% of the reserve price set for that auction unless the plaintiff and defendant otherwise agree;
(j) execute the contract for sale and if the plaintiff and defendant fail to agree on the terms of the contract for sale, the parties must execute the contract for sale in the terms recommended by the Solicitor;
(k) co-operate in every way with the agent in relation to inspections by prospective purchasers of the Property at all times requested by the agent and ensure that the Property is in a neat and clean condition; and
(I) execute all other documents necessary to complete the sale within the time required by the contract for sale to ensure that the purchasers do not have a right to terminate or rescind due to failure to do so.
4. On settlement of the sale of the Property, the parties are to authorise and direct the Solicitor to disburse the proceeds of sale (including the deposit) in the following order and priority:
(a) to the agent in payment of the agent's commission, fees and expenses;
(b) to the Solicitor for their fees in relation to the sale;
(c) to the extent that any party has paid to the agent any marketing or other costs requested by the agent, in reimbursement of the amount of that payment;
(d) in respect of any necessary adjustment for rates;
(e) as to the balance, and subject to order 5 below, in the proportions one third to the defendant and two thirds to the plaintiff.
- As to the amount currently held by the managing agent of the Property, one third is to be paid to the defendant and two thirds to the plaintiff subject to adjustments in respect of:
(a) $6,000 already paid to the plaintiff; and
(b) amounts paid by the plaintiff in respect of insurance for the Property.
- Any payment to a party of the balance of the sale proceeds in accordance with order 2(e) above is to be subject to an adjustment in respect of the amount of $17,738.83 plus interest pursuant to s 101 of the Civil Procedure Act 2005 from 13 April 2023 to the date of the payment.
- ORDER that each party to these proceedings otherwise pay or bear his or her own costs of the proceedings.
- RESERVE to the parties in the working out of these orders (in the event of any dispute about their operation) liberty to apply for orders (under the inherent jurisdiction of the Court, or Part 27 of the Uniform Civil Procedure Rules 2005 NSW or otherwise) for a judicial sale of the Property.
- ORDER that these orders be entered forthwith.
By notice of appeal filed 4 April 2025, the appellant relied on the following nine grounds of appeal:
Regardless of which party's version of events was accepted, His Honour should have held that the whole of the defendant’s legal interest in the property was held subject to a resulting trust in favour of the plaintiff.
His Honour erred by failing to make a finding as to whether the constructive trust found at [125] of the reasons dated 25 February 2025 (Reasons) was a constructive trust based upon a failed joint endeavour or a constructive trust based upon a common intention.
To the extent that the constructive trust found at [125] was one based upon a common intention that the defendant's contributions to the “Scott Street property” be recognised, his Honour erred by finding that either the plaintiff or defendant had that intention in circumstances where:
(a) neither party claimed in their evidence to have had such an intention; (b) such an intention would have been inconsistent with both the plaintiff's claimed intention (that the defendant was allowed to go on title in anticipation of marriage) and the defendant's claimed intention (that she was allowed to go on title in anticipation of assisting with the development of the property); and (c) it was never put to the plaintiff that he had such an intention.
(a) finding such a constructive trust in circumstances where neither party alleged that there was any such joint endeavour; (b) failing to take into account contributions by the plaintiff to the acquisition and development of the Cumberland Reach property; (c) failing to make any findings as to which of the payments alleged by the defendant were contributions to any joint endeavour; (d) failing to make any findings as to which of the financial contributions claimed by the defendant to that joint endeavour had already been the subject of a reimbursement or other payment to the defendant by the plaintiff; and (e) failing to make any findings as to whether the financial or non-financial contributions by the defendant had resulted in any benefit to the plaintiff.
(a) findings as to whether the defendant was the person who had created the false credit; and (b) card statements relied upon by the defendant.
The appellant, who was represented by Mr Phillips of counsel, made oral submissions closely conforming to his written submissions which I will address when dealing with each of the grounds of appeal.
The respondent, who was self-represented before the primary judge and in this Court, candidly explained that she had used an artificial intelligence program (in conjunction with some or all of the material in the appeal books) to prepare her written submissions and certain “slides” which she read to the Court as part of her oral address. This gave rise to a number of difficulties. The first was that a large number of authorities were referred to by the respondent. Most had little, if anything, to do with the issues in this case. One authority, at least, was an hallucination. Secondly, the written and oral submissions made by the respondent travelled well outside the issues raised by the appeal and addressed, for example, claims that if they were available to be made at all, should have been made by cross-appeal. It would be unfair to the appellant to act upon claims forming no part of the appeal. Thirdly, whilst a sincere attempt was made by the respondent to address the legal issues as she understood them, her submissions were of no real assistance to the Court. The respondent’s written and oral submissions were a cogent demonstration that the use of artificial intelligence by non-legally trained users is likely to add to the cost and complexity of legal proceedings without appreciable benefit. It may be that more intrusive case management techniques will need to be employed in future to seek to prevent self-represented litigants from unfairly increasing the costs and complexity of proceedings by the use of artificial intelligence. Since drafting the above, I have read the additional remarks of the Chief Justice about the use of Generative AI in the preparation of submissions, with which I agree.
In essence, the appellant’s ground 1 asserted that the primary judge failed adequately to address the appellant’s case that a “presumption of resulting trust” (focusing on the appellant’s payment of the purchase price of the Scott Street property) or a “presumption of advancement” were determinative of the issues. The appellant asserted that the primary judge should have accepted his case below that resulting trust was the correct prism through which to view the parties’ relationship and that the presumption of advancement was here rebutted as it was “an advancement conditional upon the marriage [of the parties] actually taking place”, or in the alternative, conditional upon the renovation of the Scott Street property actually occurring.
I am unable to agree with the appellant’s contentions.
As the authors of Jacobs*’* Law of Trusts in Australia (5th ed, Butterworths) explain (at [1301]), “[t]he constructive trust differs from the resulting or implied trust in that, although a resulting or implied trust also arises by operation of law, the courts presume that a trust was actually intended and in the face of evidence to the contrary, may conclude that the presumption has been rebutted”.
At the outset, the primary judge rejected the appellant’s assertion that this case could be determined by the application of a presumption of resulting trust and a consequential determination of a presumption of advancement:
[16] In my assessment the parties’ dispute is not to be determined by reference either to a “presumption of advancement” (depending perhaps on what one makes of a de facto relationship) or a “presumption of resulting trust” (focusing on the plaintiff’s payment of the purchase price of the Scott Street property) because the particular facts of the case govern its outcome. If otherwise applicable they must be taken to have been rebutted. A determination of the facts depends upon an assessment of the whole of the evidence, including questions of credit affecting fact-finding and inferences to be drawn from the evidence.
His Honour was correct to so conclude. This finding is sufficient to reject ground 1. The primary judge found, for reasons he explained at some length, that a presumption of resulting trust did not govern this case based upon the findings of fact his Honour made. If the presumption of resulting trust were to be applied, the primary judge found, on the basis of those facts, that the presumption was rebutted. Although the appellant, somewhat faintly, complained about the absence of relevant findings about an intention immediately to convey a beneficial interest in the Scott Street property to the respondent, the primary judge found at [106] (set out at [36] above), that the purchase of the Scott Street property in joint names of the parties was with the intention that, from the time of purchase, they both be beneficially entitled to the property as co-owners.
As to the appellant’s complaint that any beneficial interest in the Scott Street property was conditional upon a marriage taking place and/or the property being renovated by the parties, the primary judge’s findings about “a critical piece of the appellant’s evidence” must be examined:
“19 I was intending to purchase Scott Street in my sole name. However, in about early June 2022 the defendant said to me ‘”If you really want to marry me you should put my name on the contract for purchase of 9 Scott Street. My name should be on the title with yours.”
20 I replied to her, “l would not feel comfortable doing that.”
21 The defendant said, “You need to show me you love me, you need to show me that you intend to be my husband. I am not going anywhere. We will be together forever. What is the problem?”
22 I recall I did not respond to the defendant. I did not feel comfortable with the situation.
23 The defendant continued to pressure me by saying things like, “Should I call the conveyancer?”; “Have you called the conveyancer?”; and “What is your problem, don’t you love me?”
24 In or about mid-June I sat down with the defendant and we had a conversation to the effect:
I said: “Do you love me?”
She said: “You know I do.”
l said: “You want me to put your name on the title to the Scott Street property.”
She said: “I am going to be your wife, we are going to spend the rest of our lives together, till death do we part’’
I said: “You need to stop with all these games, I feel like you are hiding from me, lying to me, you spend all your time on social media I need you to start being straight with me, talk to me. You say you want to be my wife and that you want to spend the rest of your life with me. l want to believe what you are telling me. You say I need to show you that I love you but it is you who needs to show me. I will tell the conveyancer to put you on the title but it is only on the basis that you are going to be my wife and I need you to step up, I need you to show me the respect I deserve as your husband, the same respect I show you. You need to give back to this relationship not just take, take, take.”
She said: “I will.”
25 Following that conversation with the defendant I telephoned the conveyancer and instructed her to put the defendant’s name as joint purchaser with me on the contract.”
The primary judge found that the words the appellant here attributed to himself, “only on the basis that you are going to be my wife”, lacked the certainty one might reasonably expect of a condition of marriage imposed on a specific property transaction. I respectfully agree with the primary judge. As his Honour pointed out, the evidence nowhere disclosed a concrete demand, a plan for a marriage ceremony or a sense of necessity attaching to a marriage ceremony. Neither party appeared to have felt the need to set a date for a formal ceremony. I agree with the primary judge that the words the appellant attributed to himself are not far removed from a simple affirmation of the parties’ ongoing de facto relationship. The alternative complaint, that any beneficial interest in the Scott Street property was conditional upon the property being renovated by the parties, finds no support in the appellant’s account.
The primary judge recited, and resolved, conflicting evidence given about this conversation by the respondent:
“[129] In or about mid-June 2022, I sat down with Michael, and we had a further conversation about Scott Street. During that conversation, words were said to the following effect;
Michael: I want to flip for profit and you to help me manage and run this renovation.
Lila: It would be idiotic of me to continue to build your worth without being formally recognised as promised on the Cumberland Reach property. …
[130] Michael then walked away.
[131] When we decided to purchase Scott Street the contract for purchase was in Michael’s name solely. I didn’t think of broaching the topic of joint tenancy in lieu of his promise to put Cumberland Reach in both names. However, I refused to do any further renovation work on his properties unless I was included on title.
[132] Subsequently, in fear of him not talking to me, I sent Michael the following text;
I understood from the outset that arrangement was that if I paid for the renovations, living expenses and managed the renovation project(s), I would be included on title for Cumberland Reach and any future renovation projects so that I could be compensated for both my financial and non-financial contributions.
[133] Michael approached me and replied with words to the following effect:
Michael: “You know I can’t put your name on it because you have dependants.”
Lila: “This one didn’t need financing, you’re always telling me to stick to ‘our agreement’ but you’re not’
[134] Not long after, I called my son. I was on the balcony and Michael was around the other side of the balcony which was 10m away. I was making arrangements for a tenant at a property owned by my son to vacate in so I could move into that property as I was upset that Michael was not adhering to our agreement. I was going to walk out on the relationship at that point.
[135] In June 2022, I sent a further text message to Michael stating:
‘it would be idiotic to continue to build your wealth, and not be formally recognised for the work being put in as previously agreed on the Cumberland Reach property. I won’t continue to work in the capacity of building up your property portfolio and business ventures, and expected to work 7-8hrs a day and run a household, and not be formally recognised! As such I need to look out for myself now, and work on my own personal endeavours”.
[136] The offer was accepted and Michael approached later that day or the following day and said words to the following effect:
Michael: “If I put your name on it, will you agree to renovate”
Lila: “Absolutely, I will, thank you for formally recognising my contributions”
Elated I gave him a big hug, he called the Michelle from Crown Legal who was the conveyancer.
[137] Following this discussion, Michael, myself and the conveyancer had a three-way phone conference.
[138] During this conference, it was discussed that a Deed of Recission would be entered into for the original contract on the Scott Street property and the conveyancer would instruct Lance of LJ Hooker Fraser Coast to reissue the contract in both names. I attended with Michael signed the document in front of Angus Bennet, a mutual acquaintance, and I emailed the contract and Deed of Recission back to Crown Legal.
[139] I felt that my work had finally been recognised. I was elated, gave Michael a big hug, and thanked him. I did not move out and we started planning the renovations of Scott St, which involved me liaising with contractors to get costs for works such as new VJ board to exterior and repair of collapsing pergola, plumbing for a bathroom reno, chippy to repair wood rot in bathroom floor and a painters quoting.
[140] I arranged another inspection through Lance to devise the best approach to renovating this property and start feasibility studies of what would be needed to bring up to scratch.
[141] Shortly after, Michael told the neighbours about the purchase and their friends knocked on our front door begging us to rent out the property as they and their children had been living in tents due to the rental shortage.
[142] Michael told me that he wanted to take this option instead of renovating.
[143] At Michael’s request I arranged for Twenty Twenty realty to go through the usual tenancy checks and manage the property.”
The primary judge found that the respondent’s version of this conversation is more probably correct than the appellant’s version: at [108]. The sole factual challenge made by the appellant to the primary judge’s reasons was limited to this conclusion in paragraph [108]. I am unable to accept that this challenge has been made good. The appellant’s version was based upon a reconstruction of events through a prism of romantic attachment, and the loss of it. I agree with the primary judge that the respondent’s version fitted more realistically into the broader context, including the economic necessity to provide for herself and her younger children in a committed relationship for which she had sacrificed her own interests in reliance upon the appellant’s promises of shared mutual benefit. Although the appellant challenged his Honour’s findings about this conversation, no error, let alone error of the kind required by Fox v Percy (2003) 214 CLR 118; [2003] HCA 22 and Lee v Lee (2019) 266 CLR 129; [2019] HCA 28, was shown.
The primary judge found that the appellant had not established a basis upon which a presumption of resulting trust could operate, nor that, if such a presumption were applied, it had not been rebutted. As his Honour explained, the appellant and the respondent bought the Scott Street property as joint tenants, each recognising the respondent’s significant financial and non-financial contributions to the renovation of the Cumberland Reach property and with the common intention that the Scott Street property would be the subject of a like restoration project, the benefits of which would be mutually enjoyed by the parties. Both parties wanted a stable, productive home environment and looked to property renovation projects as a means of mutual profit.
Ground 1, which asserted that the primary judge should have found that the respondent’s legal interest in the property was held subject to a resulting trust in favour of the appellant, must be rejected.
Grounds 2 and 3 are predicated in an assertion that the primary judge found that there was a constructive trust based upon a common intention. As I have earlier explained, the primary judge found that the facts established a constructive trust based on a failed joint endeavour, rather than a common intention.
Although the primary judge occasionally referred to the “common intention” of the parties in purchasing the Scott Street property (see eg J[106]), the primary judge’s analysis clearly relies upon a joint endeavour constructive trust. At J[112]-[116], the primary judge discussed the principles of a joint endeavour constructive trust as discussed in Muschinski v Dodds and Baumgartner. At [116], whilst acknowledging that the facts of those two cases are not “comparable in all respects with the present case” and without suggesting that the “equitable principles of the character discussed in those cases can be applied as rules”, the primary judge nevertheless went on to note that “equity generally operates upon an entitlement to prevent a person from asserting or exercising a legal right in circumstances where the particular assertion or exercise of it would constitute unconscionable conduct”. His Honour then applied the elements required to be found in a joint endeavour constructive trust to the facts. These elements include the existence of a joint endeavour (see J[117]), the breakdown of the joint endeavour without attributable blame (see J[118]), and what is required to do equity between the parties (see J[119]-[125]).
In oral submissions, Mr Phillips, on behalf of the appellant essentially accepted that the better characterisation of the primary judge’s reasons was that his Honour found a constructive trust based on a failed joint endeavour:
On further reading of the decision, in my submission, the better characterisation of the reasons is probably that it's a failed joint endeavour constructive trust, notwithstanding a few references to there being a common intention. If that is the way that your Honours also read the reasons then grounds 2 and 3 fall away.
Grounds 4 and 5 address similar complaints about aspects of the primary judge’s reasons and may conveniently be dealt with together. Each ground is predicated on a finding of constructive trust based upon the failed joint endeavour.
Ground 4 complained that his Honour erred by failing to make a finding as to whether the joint endeavour was for the acquisition and development of the Scott Street property alone or the acquisition and development of both the Scott Street property and the Cumberland Reach property. I am unable to agree.
It appears to be common ground that the acquisition and development of the Scott Street property formed part of the failed joint venture. As to Cumberland Reach (and Ann Street) forming part of the failed joint venture, at the risk of repetition, the primary judge found that the respondent’s contribution to the joint endeavour and the “material welfare of both parties” was substantial, both financially and non-financially. That finding was predicated upon contributions to the expense of renovation works to the Cumberland Reach property and expenses associated with the Ann Street property and to household expenses while the parties were living at the Cumberland Reach and Ann Street properties. His Honour found that the respondent resigned from her employment as a local government Building Services Officer to manage and carry out renovation works to the Cumberland Reach property. The respondent’s non-financial contribution went well beyond household chores. She personally performed work that required skill and care in management of renovation work and in location of the Ann Street and Scott Street properties. The respondent’s financial contributions devoted to renovation of the Cumberland Reach property was a substantial proportion of the funds available to her.
I would reject ground 4.
Ground 5 is predicated on an assertion that the primary judge found a constructive trust based upon a failed joint endeavour that only included the acquisition and development of the “Scott Street property”. As I have said, that assertion is unsound.
Ground 5 should be rejected.
Ground 6 is predicated on the existence of a joint endeavour constructive trust encompassing the Cumberland Reach property and the Scott Street property. Ground 6(a) complained that “neither party alleged that there was any such joint endeavour”, as the one found by the primary judge. This ground was not developed in writing or orally. The appellant below claimed, in the alternative, a declaration that the respondent's interest in the Scott Street property is held “subject to a … constructive trust in favour of the plaintiff”. On all of the evidence, the primary judge was entitled to impose a constructive trust based on a failed joint endeavour. The topic was squarely raised by the way the case was conducted. There was no unfairness to the appellant in his Honour reaching the conclusion that he did. To the extent ground 6(a) was pressed, I would reject it.
Ground 6(b) asserted that the primary judge failed to take into account contributions made by the appellant to the acquisition and development of the Cumberland Reach property. I do not agree. His Honour said (at [117]) that “the parties acquired the Scott Street property for a common purpose (as a joint investment property, anticipating renovation and sale) and, allowing for past, present and prospective contributions to their common wealth, intending that their common purpose would be pursued, indefinitely, for their joint benefit.” His Honour ultimately concluded:
[122] In assessing the parties' respective contributions to the common wealth, I am conscious of a need to give effect to practical considerations, rather than to pursue complicated factual inquiries "which may result in relatively insignificant differences in contributions and consequential beneficial interest" (Baumgartner v Baumgartner (1987) 164 CLR 137 at 150) and, in the formulation of an equitable remedy, to endeavour to do what is "practically just", accepting that once the Court has determined upon the existence of a necessary equity to attract relief, the moulding of relief may produce a final result which, after balancing competing interests, may not exactly represent what any party may have wished: Vadasz v Pioneer Concrete (SA) Pty Ltd (1995) 184 CLR 102 at 113-114; Bridgewater v Leahy (1998) 194 CLR 457 at 493-494; Delaforce v Simpson-Cook (2010) 78 NSWLR 483 at [57] and [60]-[61]. (emphasis added)
The primary judge sufficiently took into account contributions made by the appellant to the acquisition and development of the Cumberland Reach property. I would reject ground 6(b).
Ground 6(c) complained of the primary judge's failure to make findings about which particular payments made by the respondent were contributions to the joint endeavour. I reject this complaint. The primary judge found that “it is common ground that she did make a substantial financial contribution to renovation of the Cumberland Reach property” (“in excess of $50,000 or thereabouts”) (J[121]), whether or not “the extent of that contribution cannot reliably be determined from her records” (J[100]). The primary judge also found, on the basis of the evidence including his Honour’s credibility findings, that after the appellant terminated their relationship, the appellant refused to pay or allow the respondent any compensation for the work she had done, and the expenses she had incurred, in renovation of the Cumberland Reach property. I would reject ground 6(c).
Ground 6(d) complained that the primary judge failed to make findings about contributions claimed by the respondent to the joint endeavour which allegedly had already been the subject of a reimbursement to the respondent by the appellant. In writing, the appellant asserted that a “reimbursement” of $10,300 had been made and it was suggested that the reimbursement may have been overlooked. In oral address, Mr Phillips for the appellant stated that there had been a number of alleged reimbursements and almost all the amounts spent by the respondent were in dispute.
This complaint must be rejected. As I have explained, the primary judge made a factual finding that the respondent’s financial contribution to the renovation of the Cumberland Reach property was “in excess of $50,000 or thereabouts” (J[121]). That finding was not challenged as part of the appellant's UCPR r 51.36 statement. Mr Phillips accepted that he had not challenged that finding of fact. Ground 6(d) should be rejected.
Ground 6(e) complained about an alleged failure to make findings about whether the financial or non-financial contributions made by the respondent had resulted in any financial benefit to the appellant. The starting point is that it must be shown that “the benefit of money or other property contributed by one party [was] on the basis and for the purposes of the [joint] relationship or endeavour” (Muschinski v Dodds at 620). What must be demonstrated is both the existence of a joint relationship or endeavour, and that the parties made contributions on the basis and for the purpose of that joint relationship or endeavour. Non-financial contributions may affect the parties’ beneficial interests in property under a joint endeavour constructive trust: Baumgartner at 149.
The common benefit, to the extent it is required to be shown, does not need to be material in the sense of a financial benefit. In [2024] NSWCA 258, Bell CJ (Gleeson JA and Stern JA agreeing) said:
[45] … Apart from needing to be understood in the specific context of the facts of that case and in light of observations in Baumgartner to which Campbell J referred, as already explained, in the present case, there was “common benefit” in the Joint Endeavour for both Ravina and Ronnie and Marie. There is no reason in principle why the “common benefit”, to the extent that it may be required to be shown (and Muschinski does not suggest that this is a requirement for the imposition of a remedial constructive trust), must be material in the sense of a financial benefit.
In the present case, the primary judge identified the contributions made by the respondent to the joint endeavour. Those contributions were specifically identified by the primary judge and included her financial contributions to the renovation of the Cumberland Reach property, household expenses and skill and care in the management of renovation work of the Cumberland Reach, the Ann Street and Scott Street properties. The primary judge was not required to go further and identify some particular financial benefit to the appellant. I would reject ground 6(e).
Ground 6 should be rejected.
By ground 7, the appellant asserted that the conclusion of the primary judge that the parties held the Scott Street property in proportions of two-thirds and one-third was incorrect. I would reject this ground.
The essence of this ground is that the respondent made “no contributions” to the acquisition of the Scott Street property. At the risk of repetition, the primary judge’s conclusion was informed by the respondent’s financial and non-financial contribution to the joint endeavour. As his Honour found, “that she made both a financial and a non-financial contribution to the material welfare of both parties is not in dispute”. The respondent was named as a co-owner of Scott Street deliberately and mindful of her expectation of co-ownership arising from her proven contribution to renovation of the Cumberland Reach property and a common expectation that the Scott Street property would be the subject of a like restoration project. In so doing, the primary judge clearly took into account the fact that it was the appellant who contributed the whole of the purchase price. That was the principal basis of the primary judge’s finding that it would be unconscionable for the respondent to insist upon retaining beneficial ownership of a full one half share of the Scott Street property.
The appellant’s fall-back submission, that he should have been awarded a 95% share of the Scott Street property should likewise be rejected. The primary judge considered, in some detail, the entire nature of the joint endeavour between the parties and identified how it was, in context, that from the outset the respondent was named as a co-owner, not merely a nominee of the appellant, of the Scott Street property. Having explained the facts relevant to this case, the primary judge’s orders were moulded to the justice of the case in an orthodox manner.
I would reject ground 7.
Ground 8 complained that the primary judge erred by making the factual findings at [108]-[110] without having regard to certain of the contemporaneous written records relied upon by the appellant. During oral submissions, counsel for the appellant accepted that this complaint was one that his Honour’s finding was glaringly improbable of the kind identified in Fox v Percy.
I would reject ground 8. First, the primary judge saw the appellant and the respondent give evidence. His Honour was in a much better position than this Court to make findings about the context in which relevant text messages were exchanged. As I have discussed at [33], the primary judge concluded that the respondent was a “better historian” and that her version of the critical conversation (set out when addressing ground 1 above) was more likely to be correct. That finding has not been shown to be glaringly improbable.
Secondly, a close examination of the text messages relied upon by the appellant actually supports the conclusions reached by the primary judge. In particular, on 20 June 2022, prior to the execution of the relevant contract for sale of the Scott Street property, the respondent said to the appellant:
I know my worth and if you don’t see it by now, and can’t offer me guaranteed security for the children and I, in the event our relationship ends tomorrow and you threaten to kick us out again. As much as you then turn around and say you love me/us and want to spend forever together. I need to focus on that. I’ve worked as a team to build a future together and you’ve implied I’d be on the street. It’s a very unsettling situation (emphasis added).
It is tolerably clear that the respondent here is referring to contributions made by her in the past to the joint endeavour in a way consistent with the factual findings at [108]-[110] made by the primary judge.
Ground 8 should be dismissed.
Ground 9 complained that his Honour erred in assessing the respondent’s credit at [97]-[98] by failing to make findings about whether the respondent was the person who had created false credit card statements.
This complaint is without substance. The primary judge dealt with this issue and made a serious credit finding about certain falsified credit card entries, which amounted to approximately $5,000 being overclaimed by the respondent as contributions to the joint endeavour:
[97] I accept that some of the credit card statements relied upon by the defendant in these proceedings include patently falsified entries and, because the defendant professes to have no explanation for them, they bear heavily upon her credit, at least in relation to quantification of her financial contribution to renovation of the Cumberland Reach property and the payment of household expenses.
For the reasons given, I would dismiss the appeal with costs.
At the hearing, it was common ground that the Scott Street property had been sold and settlement was to occur on 11 August 2025. Contingent orders were made for the payment into Court of the net settlement sum. As this judgment is being published in advance of 11 August, and the appeal has been dismissed, the orders made by Linsday J on 10 March 2025 provide for payments to be made from that settlement sum, including disbursements.
McHUGH JA: I agree with the orders proposed by Payne JA, and with his Honour’s reasons. I also agree with the Chief Justice’s observations concerning the use of Generative AI in the preparation of submissions.
**********
May
Costaras [2025] NSWCA 178
(1987) 164 CLR 137
(2003) 214 CLR 118
(2019) 266 CLR 129
(1985) 160 CLR 583
(1995) 184 CLR 102
(1998) 194 CLR 457
(2010) 78 NSWLR 483
Appeal dismissed. The appellant was ordered to pay the respondent’s costs. The primary judge’s declaration of beneficial interests in the proportions of two-thirds for the appellant and one-third for the respondent remained in place.
PAYNE JA: The appellant, Michael May, and the respondent, Lila Costaras, are proprietors, as joint tenants, of a property in Scott Street in Maryborough in Queensland. In 2022, the appellant commenced proceedings in the NSW Supreme Court seeking to establish that the respondent held her legal interest in the property on trust solely for him.
The primary judge, Lindsay J, found that the parties held the Scott Street property on a joint endeavour constructive trust of the kind recognised in Muschinski v Dodds (1985) 160 CLR 583; [1985] HCA 78 and Baumgartner v Baumgartner (1987) 164 CLR 137; [1987] HCA 59. His Honour made orders declaring that the beneficial interest of the property was to be divided in the proportion of two thirds for the appellant and one third for the respondent: May v Costaras [2025] NSWSC 90.
For the reasons which follow, the primary judge did not err in so concluding and I would dismiss the appeal.
The primary judge found the fact that the Scott Street property was purchased in the names of both parties as joint tenants was consistent with the parties’ relationship being “in the nature of a marriage”: