[1] Starting in the 1960s, provincial governments began to apply their child welfare legislation to children living in First Nations communities. This resulted in the massive apprehension and removal of Indigenous children from their communities, as well as their placement and adoption in non-Indigenous families. The Sixties’ Scoop, as this practice came to be known, had a profound detrimental effect on the children affected and their communities. For many of them, being raised in a non-Indigenous family estranged them from their Indigenous culture and identity. Many have experienced significant difficulties when they became conscious of their Indigenous identity and faced major hurdles when attempting to reconnect with their communities. The Sixties’ Scoop broke the social fabric of the communities whose children were removed. Its consequences are often compared to those of the residential schools.

[2] Canada funded the Sixties’ Scoop—at least with respect to First Nations (or “status Indian”) children. Provinces were initially reluctant to apply their child welfare legislation to them, because they viewed this as a federal responsibility. Canada therefore undertook to reimburse the provinces for the expenses incurred for extending their services to First Nations communities (or “reserves”).

[3] In 2017, the Ontario Superior Court of Justice held that Canada breached its duty of care towards First Nations children by failing to take measures aimed at facilitating the transmission of Indigenous culture and identity in children who were placed in foster families or adopted pursuant to provincial child welfare legislation. This prompted Canada to settle several class actions related to the Sixties’ Scoop. The settlement agreement concluded in 2018 provided that compensation would be paid to First Nations and Inuit survivors, but not to other Indigenous persons, in particular those who identify as Métis or non-status Indians.

[4] The present class action was brought on behalf of survivors excluded from the 2018 settlement. The parties have not been able to reach an agreement. They brought motions for summary judgment addressing what seem to be the main issues in dispute. Canada is asking the Court to dismiss the claims for monetary relief in respect of all members of the class, because the action would have been brought outside the six-year limitation period. The Plaintiffs are asking the Court to declare that Canada had a duty of care or a fiduciary duty towards Métis and non‑status Indian survivors of the Sixties’ Scoop.

[5] I am dismissing Canada’s motion. The limitation period begins to run only when a plaintiff reasonably ought to have discovered the facts on which their claim is based. The circumstances in which the survivors learned about their Indigenous identity and came to appreciate the harm they suffered vary widely. Contrary to Canada’s submissions, these individual circumstances cannot be disregarded when deciding whether each survivor’s claim is time-barred, which means that limitations is not an issue that can be resolved on a class-wide basis.

[6] I am granting the Plaintiffs’ motion, but only with respect to a subset of the class, namely, survivors who were placed or adopted through Saskatchewan’s Adopt Indian Métis [AIM] program. Canada had a duty of care towards these children because it funded a specific program that would foreseeably sever their connection with Indigenous culture and identity.

[7] Canada, however, did not have a duty of care towards the remainder of the class. Except in relation to the AIM program and other minor exceptions, Canada never funded the application of provincial child welfare legislation to Métis and non-status Indians. This was a core policy decision that is immune from liability. Moreover, proximity is lacking, as the provinces applied their laws to these children on their own initiative, without any federal involvement. Canada cannot have a duty of care with respect to the manner in which the provinces apply their own legislation. The historical relationship between Canada and Indigenous peoples cannot, without more, ground a duty of care. No fiduciary duty arises in this case.

[8] This result must be understood in light of the fundamental difference between the situation of First Nations and Inuit survivors of the Sixties’ Scoop, on the one hand, and those who identify as Métis and non-status Indians, on the other hand. Even though the effects of removal and adoption of Indigenous children were likely the same regardless of formal status, the role Canada played with respect to each group was fundamentally different. The provinces generally refused to provide child welfare services in First Nations communities until Canada agreed to reimburse the costs. Thus, Canada’s funding was critical in unleashing the Sixties’ Scoop in First Nations communities. In contrast, with the exception of AIM, Canada never funded child welfare services for Métis and non-status Indian children and their removal and adoption took place without federal intervention.

II. Background

[9] To understand my decision, it is obviously necessary to provide a short description of the Sixties’ Scoop. It is also necessary to explain the involvement of the federal and provincial governments and how the federal government’s involvement differed with respect to status Indians and Métis and non-status Indians. To accomplish this, one must have a grasp of the categories the Canadian state deployed to classify the Indigenous population. It is to this issue that I turn first.

A. State-Created Identity Categories

[10] At the heart of this case lies the use of state-created categories to describe Indigenous peoples and subject them to different policies. It is important to clarify the origin and meaning of these categories at the outset.

[11] When Canada became a federal country in 1867, Euro-Canadians commonly used the term “Indian” to describe Indigenous persons. For example, section 91(24) of the Constitution Act, 1867 empowers Parliament to make laws regarding “Indians.” A few years later, Parliament adopted the Indian Act, which defined a category of persons having legal status as “Indians” or, in short, “status Indians.” Nowadays, Indigenous peoples governed by the Indian Act are more commonly known as First Nations.

[12] Not everyone who claimed Indigenous ancestry or identity was entitled to Indian status. Parliament explicitly excluded the Métis of Western Canada from Indian status and dealt with their claims through other means. Indian status was defined in a manner that furthered the government’s policy of gradual assimilation. In particular, the Indian Act deprived Indigenous women who married non-Indigenous men of their Indian status. In addition, many Indigenous individuals or communities were inadvertently overlooked when treaties were signed and membership lists of First Nations were compiled. These features of the Indian Act and its administration thus gave rise to a population of persons associated with First Nations but who did not hold status and who came to be known as “non-status Indians.” Lastly, when the government came into closer contact with the Inuit, it chose not to bring them under the Indian Act.

[13] The exclusion of many Indigenous persons and groups from the Indian Act gave rise to a controversy regarding the scope of section 91(24) of the Constitution Act, 1867: was Parliament’s jurisdiction limited to those to whom it recognized Indian status? In 1939, the Supreme Court stated that the concept of “Indian” in section 91(24) was broader than the category of persons having status under the Indian Act. It decided that the Inuit, although not entitled to registration under the Indian Act, were nevertheless “Indians” for the purposes of section 91(24): Reference as to Whether the Term “Indian” in Head 24 of Section 91 of the British North America Act, 1867, Includes Eskimo Inhabitants of the Province of Quebec, [1939] SCR 104. As a result, it became clear that Indian status was not an exhaustive definition of Indigenous identity for constitutional purposes, although whether the Métis were included in section 91(24) remained unclear.

[14] When the constitution was “patriated” forty years later, section 35 of the Constitution Act, 1982 recognized and affirmed the aboriginal and treaty rights of “the Indian, Inuit and Métis peoples of Canada”, grouped together under the more generic description of “aboriginal peoples.” The Supreme Court later clarified that the term “Indian” in section 91(24) of the Constitution Act, 1867 encompasses all aboriginal or Indigenous peoples, including the Métis and non-status Indians: Daniels v Canada (Indian Affairs and Northern Development), 2016 SCC 12, [2016] 1 SCR 99 [Daniels].

[15] While the term “Indian” is the result of a geographical error and is now usually replaced by “First Nation,” I will nevertheless use the term “status Indian” in these reasons when referring to persons entitled to status or registration pursuant to the Indian Act. Like the Plaintiffs, I will also use the terminology of “Métis and non-status Indian” as a shorthand for all Indigenous persons who are neither status Indians nor Inuit.

B. Indigenous Peoples and Child Welfare

[16] The identity categories outlined above are critical in understanding the manner in which the federal and provincial governments assumed responsibility for the provision of public services to Indigenous peoples, in particular child welfare.

[17] Sections 91–95 of the Constitution Act, 1867 allocate jurisdiction to enact legislation as between Parliament and the provincial legislatures. The usual assumption is that the distribution of executive power, including the financial responsibility for the provision of public services, mirrors that of legislative power. For this reason, the provinces have frequently taken the position that they do not have any constitutional responsibility to provide services to status Indians, especially those who reside on reserves.

[18] In the mid-20^th century, the federal government sought to integrate Indigenous peoples into the general citizenry and make them eligible for the same public services as other citizens. To that end, it invited provinces to extend the public services they were already providing to other citizens, including child welfare services, to status Indians residing on reserves. In 1951, this policy was reflected in the enactment of what is now section 88 of the Indian Act, which states the principle that, subject to certain exceptions, provincial laws apply to status Indians.

[19] Given that they viewed the provision of services to status Indians as a federal responsibility, the provinces were reluctant to respond to the federal invitation to provide services to status Indians residing on reserves, unless they received financial compensation. Thus, to achieve its policy goal, the federal government had to reimburse the provinces for the costs of providing these services. With respect to child welfare more specifically, it entered into formal agreements or less formal arrangements, starting in the 1960s, whereby it would reimburse the provinces for the provision of such services to status Indians residing on reserves. These agreements provided that the services must comply with provincial legislation. The main aspect of child welfare services relevant to the present action is the apprehension of children whose well-being is compromised in their birth families and their placement in foster families or for adoption. This aspect of child welfare services is often called child protection.

C. The Sixties’ Scoop

[20] When the provinces began in earnest to apply their child welfare laws in First Nations communities in the 1960s, this resulted in the massive removal of First Nations children from their communities and their placement in foster care or for adoption in non-Indigenous families. This became known as the Sixties’ Scoop.

[21] In its 2015 report, the Truth and Reconciliation Commission summarized the Sixties’ Scoop as follows:

The provincial social workers assigned to reserves assessed child safety and welfare by mainstream cultural standards. They received little or no training in Aboriginal culture. They were not trained to recognize problems rooted in generations of trauma related to the residential schools. Instead, they passed judgment on what they considered bad or neglectful parenting. As a result, beginning in the 1960s, provincial child welfare workers removed thousands of children from Aboriginal communities. It has been called the “Sixties Scoop.”

Aboriginal children were placed in non-Aboriginal homes across Canada, in the United States, and even overseas, with no attempt to preserve their culture and identity. The mass adoptions continued between 1960 and 1990.

The Sixties Scoop children suffered much the same effects as children who were placed in residential schools. Aboriginal children adopted or placed with white foster parents were sometimes abused. They suffered from identity confusion, low self-esteem, addictions, lower levels of educational achievement, and unemployment. They sometimes experienced disparagement and almost always suffered from dislocation and denial of their Aboriginal identity.

(TRC, Canada’s Residential Schools: The Legacy. The Final Report of the Truth and Reconciliation Commission of Canada, Vol. 5, pp. 14–15)

[22] As this excerpt shows, descriptions of the Sixties’ Scoop usually focus on status Indians living in First Nations communities (or “reserves”). This is likely because provincial child welfare authorities did not, for the most part, serve these communities before the federal government agreed to provide funding in this regard. Nevertheless, provincial child welfare authorities also apprehended Indigenous children who did not have Indian status (such as Métis children) or children with Indian status not living on reserves. The evidence does not specifically show when such a practice began.

[23] What is not seriously in dispute is the fact that the harms associated with removal and placement in a non-Indigenous foster family do not depend on status and are substantially the same for Métis and non-status Indians as for status Indians. In this regard, the Plaintiffs brought the evidence of Dr. Raven Sinclair, who has worked extensively with Sixties’ Scoop survivors and states that the impacts are the same irrespective of identity categories.

[24] One historical document filed in evidence shows how the Métis themselves perceived these impacts. In a 1971 memorandum, the Métis Society of Saskatoon objected to the placement of Métis children in non-Indigenous foster families, apparently pursuant to the AIM program. The relevant portion of the memorandum reads as follows:

As Metis parents of Saskatoon, we are decidedly opposed to having our children separated from Metis homes and culture and being forced to live in white homes.

[…]

Specifically, we object to these white foster homes because:

1. Our Metis children are subject to discrimination, because in a white supremacy society, children of Indian blood are naturally rejected.

2. In white homes our children are not given genuine love and a feeling of being wanted.

3. Our children naturally feel more contented and happy in their own Metis culture.

4. Because of their Indianness and appearance our children can not really be accepted in the white society.

5. The white foster parents are able to terminate acceptance at any time the care of our children.

6. Consequently, we are shoved from foster home to foster home, continuously.

7. As Metis parents, we feel a sense of racial and cultural responsibility for our children.

8. We want our children to be brought up as Metis and not as middle class pseudo-whites.

9. These children belong in our Metis culture and nation.

10. We are opposed to a foster home scheme as a relocation or integration program.

11. We are opposed to the impersonal and dehumanizing institutional experience imposed on our foster children by white staff.

[25] In spite of opposition from Indigenous peoples, provincial authorities kept placing Indigenous children in non-Indigenous foster homes. The Supreme Court rejected two attempts to curtail this practice: Natural Parents v Superintendent of Child Welfare, [1976] 2 SCR 751 [Natural Parents]; Racine v Woods, [1983] 2 SCR 173 [Racine]. In the latter case, the Court held that in assessing the child’s best interests, “the significance of cultural background and heritage as opposed to bonding abates over time” (at 187). Beginning in the 1980s, further Indigenous opposition led several provinces to amend their child welfare legislation to provide that Indigenous identity and culture must be considered when assessing the best interests of an Indigenous child. Certain provinces imposed a moratorium on the adoption of Indigenous children by non-Indigenous parents. For its part, the federal government began to fund First Nation child welfare agencies, to which certain aspects of the application of provincial legislation were delegated. These developments did not reduce the overrepresentation of Indigenous children in provincial child welfare systems.

[26] In 2015, the Premier of Manitoba offered an official apology for the Sixties’ Scoop. Later the same year, the Truth and Reconciliation Commission concluded that “Canada’s child-welfare system has simply continued the assimilation that the residential school system started”: Honouring the Truth, Reconciling for the Future: Summary of the Final Report of the Truth and Reconciliation Commission of Canada (2015) at 186. The following year, the Canadian Human Rights Tribunal found that Canada’s underfunding of First Nations child welfare agencies was discriminatory: First Nations Child and Family Caring Society of Canada v Canada (Attorney General), 2016 CHRT 2 [Caring Society]. In 2019, Parliament sought to address the situation by enacting legislation recognizing Indigenous peoples’ jurisdiction to make their own laws in relation to child welfare and making cultural continuity an overarching principle in the application of child welfare laws: An Act Respecting First Nations, Inuit and Métis children, Youth and Families, SC 2019, c 24 [Bill C-92].

D. The Brown/Riddle Class Actions

[27] Beyond legislative reforms aimed at the future, providing compensation to survivors of the Sixties’ Scoop remained an issue. Several class actions were initiated for this purpose, some against Canada only and some against provincial governments as well. In one of them, the Ontario Superior Court of Justice issued a summary judgment holding that Canada breached its duty of care towards First Nations children who were removed from their families and placed with non-Indigenous adoptive or foster parents: Brown v Canada (Attorney General), 2017 ONSC 251 [Brown].

[28] In that case, Justice Belobaba found a duty of care because the federal government breached a provision of its 1965 agreement with Ontario that required consultation with each First Nation before extending provincial child welfare services to it. Had there been consultation, Justice Belobaba found that First Nations would have suggested means for maintaining the children’s connection with their communities of origin, culture and identity. In this regard, First Nations, and by extension their members, were in the position of a third party beneficiary of a contract. According to Justice Belobaba, this was an established category in which a duty of care exists, and no further analysis was necessary.

[29] Nevertheless, in the alternative, Justice Belobaba went on to consider the approach for establishing a duty of care in novel cases and determined that a new category should be recognized in the circumstances. He found that the requisite proximity derived from the “long‑standing historical and constitutional relationship” between Canada and Indigenous peoples (at paragraph 78). It was also foreseeable that Canada’s failure to act with care would cause harm to Indigenous children by jeopardizing their Indigenous identity.

[30] In contrast, Justice Belobaba found that Canada did not have a fiduciary duty towards Indigenous children, because the matter did not relate to interests in land and the situation did not meet the test for fiduciary duties in the non-Indigenous context.

[31] After Justice Belobaba’s decision, Canada and the plaintiffs in the various class actions successfully negotiated a settlement providing for a simplified claims process leading to individual compensation as well as the creation and endowment of a foundation. The various class actions were consolidated in one national class action in the Federal Court, Riddle v Canada. Both the Federal Court and the Ontario Superior Court of Justice approved the settlement: Riddle v Canada, 2018 FC 641, [2018] 4 FCR 491; Brown v Canada, 2018 ONSC 3429.

[32] One feature of the settlement lies at the root of the present case: only status Indians and Inuit are eligible for individual compensation. Objections were made to the exclusion of Métis and non-status Indians, but my colleague Justice Michel Shore nevertheless approved the settlement. At paragraph 54 of his reasons, he found that the settlement agreement was fair despite this exclusion, because the foundation was for the benefit of all survivors, federal‑provincial agreements did not cover Métis and non-status Indians, it would be difficult to determine who would be eligible for compensation and the settlement agreement preserved the rights of Métis and non-status Indians.

E. The Present Class Action

[33] The present action was instituted in 2018 on behalf of Métis and non-status Indian persons who were apprehended and placed in the care of non-Indigenous foster or adoptive parents. The representative plaintiff Shannon Varley was born in Prince Albert, Saskatchewan, and describes herself as non-status Indian. She was apprehended at birth and adopted by non‑Indigenous parents a few months later through the AIM program. The representative plaintiff Sandra Jacqueline Lukowich was also born in Prince Albert, Saskatchewan, and states that she has Métis heritage through her mother. She was apprehended during her first year of life and adopted by a non-Indigenous family shortly afterwards. During their childhood, both plaintiffs had little or no contact with their Indigenous heritage. They never received any government service aimed at facilitating their participation in their Indigenous culture or the exercise of their rights. They now feel disconnected from their Indigenous heritage.

[34] In a nutshell, the statement of claim asserts that the federal government has a fiduciary duty and a common law duty of care towards class members. In particular, the federal government was aware that Métis and non-status Indian children were at risk of being apprehended by provincial child welfare authorities and that this would lead to a loss of their culture and identity. It breached its duty of care by failing, in essence, to ensure that the provincial child welfare systems protect Indigenous culture and that Métis and non-status Indian children be provided with appropriate services.

[35] On consent, the action was certified as a class action: Varley v Canada (Attorney General), 2021 FC 671. The class is defined as follows:

All Indigenous persons, as referred to by the Supreme Court of Canada in Daniels v. Canada (Indian Affairs and Northern Development), 2016 SCC 12, at para. 6, excluding Indian persons (as defined in the Indian Act) and Inuit persons, who were removed from their homes in Canada between January 1, 1951 and December 31, 1991 and who were placed in the care of non-Indigenous foster or adoptive parents.

[36] While the parties have engaged in negotiations, they have been unable to reach a settlement. They brought motions for summary judgment to resolve the most contentious issues. Thus, the Plaintiffs brought a motion asking the Court to find that the federal government owed a duty of care and a fiduciary duty to the class members, whereas the Defendant brought a motion seeking to have the claim dismissed because it is time-barred. I address the merits of each motion in turn below.

III. Limitations

[37] The Defendant asserts that the monetary claims of all members of the class were brought out of time and are statute-barred. In its motion for summary judgment, it seeks the certification of the limitations issue as an additional common issue, and it asks the Court to dismiss the monetary claims because the applicable limitation period has expired.

[38] I am dismissing the Defendant’s motion. Both parties agree that a class member’s claim is statute-barred if that class member reasonably ought to have known the facts constituting their cause of action more than six years before the action was certified as a class action. They also agree that I can only decide the matter as a common issue if I find that the claims of all class members are statute-barred. Contrary to the Defendant’s submissions, however, this finding cannot be made from the perspective of the “average person,” in a manner that ignores each class member’s particular circumstances. The evidence shows that in this case, the personal circumstances of class members vary widely. Moreover, the evidence brought by the Defendant, consisting of media reports and other class actions, does not establish that all class members reasonably ought to have discovered their claim before May 2015. My reasons follow.

A. Legal Framework

[39] It is useful to begin by setting out the main components of the legal framework that governs the determination of the limitations issue. The parties have helpfully agreed on the basic components of this framework, which are summarized below.

(1) Indigenous Monetary Claims

[40] First, I agree with both parties that limitation periods apply to monetary claims brought by Indigenous peoples. The parties’ common position is consistent with the decisions of the Supreme Court of Canada in Canada (Attorney General) v Lameman, 2008 SCC 14 at paragraph 13, [2008] 1 SCR 372 [Lameman]; Shot Both Sides v Canada, 2024 SCC 12 at paragraph 60 [Shot Both Sides].

(2) Federal Limitation Period

[41] I also agree with the parties that the applicable limitation period is found in section 32 of the Crown Liability and Proceedings Act, RSC 1985, c C-50, which reads:

32 Except as otherwise provided in this Act or in any other Act of Parliament, the laws relating to prescription and the limitation of actions in force in a province between subject and subject apply to any proceedings by or against the Crown in respect of any cause of action arising in that province, and proceedings by or against the Crown in respect of a cause of action arising otherwise than in a province shall be taken within six years after the cause of action arose.

32 Sauf disposition contraire de la présente loi ou de toute autre loi fédérale, les règles de droit en matière de prescription qui, dans une province, régissent les rapports entre particuliers s’appliquent lors des poursuites auxquelles l’État est partie pour tout fait générateur survenu dans la province. Lorsque ce dernier survient ailleurs que dans une province, la procédure se prescrit par six ans.

[42] Section 32 creates a “federal limitation period” applying to claims against the federal Crown where the cause of action cannot be located in a single province. This means that provincial limitation periods apply only where the cause of action can be located in a single province. Section 39 of the Federal Courts Act, RSC 1985, c F-7, draws a similar distinction with respect to the limitation period regarding claims brought in the Federal Courts, whether or not the Crown is the defendant.

[43] It appears that section 32 of the Crown Liability and Proceedings Act and section 39 of the Federal Courts Act have received a broad interpretation, which favours the application of a uniform limitation period across the country: Markevich v Canada, 2003 SCC 9, [2003] 1 SCR 94 [Markevich]. For example, in Brazeau v Canada (Attorney General), 2020 ONCA 184 at paragraph 32 [Brazeau], a class action, it was held that the federal limitation period applied because the claim related to “the adoption and maintenance of a federal regulatory policy regime regarding administrative segregation that applied in all provinces.” Conversely, it was held that for a provincial limitation period to apply, “all the elements of the cause of action must have occurred in the same province”: Apotex Inc v Sanofi-Aventis, 2013 FCA 186 at paragraph 105, [2015] 2 FCR 644; see also Canada v Maritime Group (Canada) Inc, [1995] 3 FC 124 (CA); Canada (Attorney General) v St-Onge, 2024 FCA 207.

[44] A cause of action is “a set of facts that provides the basis for an action in court”: Markevich at paragraph 27. Here, the cause of action is based mainly on the federal government’s conduct with respect to class members located throughout Canada, as in Brazeau. The federal government’s alleged omissions would have pertained to the whole country, for example, failing to ensure that child welfare systems are applied in a culturally sensitive manner. As will be explained below, these omissions resulted from a policy that applied throughout the country. Even a cause of action based on a federal-provincial agreement, or the failure to enter into one, would involve facts not exclusively located in a particular province. In addition, harm is a component of the cause of action and several class members born in one province were placed in foster care or adopted in a different province. For all these reasons, it cannot be said that the cause of action arose in a single province. Thus, the federal limitation period applies.

(3) Discoverability

[45] Section 32 of the Crown Liability and Proceedings Act is silent as to when a cause of action “arises” and triggers the beginning of the limitation period. The common law rule of discoverability supplies the answer: Doig v Canada, 2011 FC 371 at paragraph 31. According to this rule, “a cause of action arises for purposes of a limitation period when the material facts on which it is based have been discovered or ought to have been discovered by the plaintiff by the exercise of reasonable diligence”: Central Trust Co v Rafuse, [1986] 2 SCR 147 at 224 [Rafuse]. More precisely, “a claim is discovered when a plaintiff has knowledge, actual or constructive, of the material facts upon which a plausible inference of liability on the defendant’s part can be drawn”: Grant Thornton LLP v New Brunswick, 2021 SCC 31 at paragraph 42, [2021] 2 SCR 704.

B. Should an Additional Common Issue be Certified?

[46] This brings me to the first disputed issue in this motion. The Defendant is asking me to certify an additional common question, namely, whether the claims for monetary relief are barred by section 32 of the Crown Liability and Proceedings Act. The Plaintiffs oppose this aspect of the motion. They say that because of their individual nature, limitations issues can never be addressed collectively in a class action, especially when discoverability is at stake.

[47] I disagree with the Plaintiffs’ blanket denial of the possibility of dealing with limitations as a common issue. I acknowledge that in the class action context, it is often said that limitations cannot be a common issue, especially when the subjective component of the discoverability rule is at stake: Smith v Inco Limited, 2011 ONCA 628 at paragraph 164 [Smith]; Levac v James, 2023 ONCA 73 at paragraph 106 [Levac]. Nevertheless, a defendant is entitled to assert limitations as a common issue where the relevant evidence applies equally to all the members of the class. See, for example, Fresco v Canadian Imperial Bank of Commerce, 2022 ONCA 115 at paragraph 18 [Fresco CA]; Levac at paragraph 107; Spina v Shoppers Drug Mart Inc, 2024 ONCA 642 at paragraphs 132–136 [Spina]; Fehr v Sun Life Assurance Company of Canada, 2024 ONCA 847 at paragraphs 84–93 [Fehr].

[48] In this case, contrary to usual practice, I am asked to rule on the merits of the issue at the same time I am asked to certify it as a common issue. As will become clear shortly, I find on the merits that the relevant facts are not the same for each class member, which precludes a class‑wide determination. In these circumstances, no useful purpose would be served by certifying limitations as a common issue.

C. Was the Cause of Action Objectively Discoverable for All Class Members?

[49] This brings me to the second issue, namely, whether the cause of action was objectively discoverable or, more accurately, whether class members had constructive knowledge of the material facts. Both parties agree that it is not enough to conclude that a majority of the members of the class ought to have discovered the claim: Smith at paragraph 164. Rather, a common answer to this issue can only be given if all members of the class ought to have discovered the claim. The parties are also in agreement that the critical date is May 2015, that is, six years before Justice Phelan’s decision certifying the class action. In other words, if I reach the conclusion that all members of the class ought to have discovered the claim before May 2015, then the claim is time-barred.

[50] The Defendant is arguing that the claim was objectively discoverable before May 2015 because some members of the class brought other class actions dealing with the same issues before 2015 or because the media attention that these class actions received would have alerted a reasonably diligent class member to the main facts constituting the cause of action.

[51] I am unable to agree with the Defendant. The main premise of its argument is that objective discoverability must be assessed in the abstract, without reference to each individual class member’s circumstances or abilities. For the reasons set forth below, whether an individual member ought to have discovered their cause of action is an inquiry that cannot be divorced from a person’s actual circumstances. In the rare cases in which discoverability was decided as a common issue, there was evidence that the circumstances of each class member were substantially similar. Such evidence is lacking in the present case. It follows that the introduction of class actions by other class members or the media attention that they attracted cannot be the basis for a finding that all members of the present class ought to have discovered their cause of action before May 2015.

(1) Objective Discoverability is Context-Dependent

[52] The basic premise of the Defendant’s submissions is that the objective (or constructive) component of the discoverability rule leaves no space for consideration of individual circumstances. To put it bluntly, if one class member ought to have discovered their claim, all class members ought to have done the same. The Defendant argues that the Court must adopt the perspective of the “reasonable average person” and disregard individual circumstances in assessing objective discoverability.

[53] I fail to see any basis, in law or in common sense, for such a proposition. The Defendant cited no authority for it. Of course, whether someone ought to have discovered something is a judgment made by an external observer and, for this reason, it is said to be objective. Nevertheless, this judgment must, at least to a certain extent, be based on the person’s circumstances, such as the person’s access to information sources, means of communication and professional advice, level of education, linguistic abilities, and so forth.

[54] Indeed, in Grant Thornton, the Supreme Court stated that “a claim is discovered when a plaintiff has knowledge, actual or constructive . . .” [Emphasis added]. It did not refer to a reasonable person’s constructive knowledge. In Rafuse, the test was described as whether the facts “ought to have been discovered by the plaintiff” [Emphasis added], not by an average person. Likewise, in Peixeiro v Haberman, [1997] 3 SCR 549 at paragraph 39, the Court emphasized “the fundamental unfairness of requiring a plaintiff to bring a cause of action before he could reasonably have discovered that he had a cause of action” [Emphasis added]. It would be equally unfair to hold that the plaintiff’s claim is time-barred because someone else, in different circumstances, could have reasonably discovered the facts. The Ontario Court of Appeal made this explicit in Longo v MacLaren Art Centre, 2014 ONCA 526 at paragraph 43, where it stated that the objective discoverability analysis “will include an analysis of not only the nature of the potential claim but also the particular circumstances of the plaintiff.”

[55] An inquiry into whether a plaintiff ought to have discovered certain facts by exercising due diligence will usually be based on the actual facts of the case, more specifically whether the plaintiff was made aware of certain facts that would have prompted a reasonable person to make further inquiries: see, for example, Canadian Natural Resources Limited v Husky Oil Operations Limited, 2020 ABCA 386 at paragraphs 34–36; Milota v Momentive Specialty Chemicals, 2020 ABCA 413 at paragraph 22. In the class action context, Justice Belobaba summarized the approach as follows in Fresco v Canadian Imperial Bank of Commerce, 2020 ONSC 6098 at paragraph 16, aff’d Fresco CA:

What the claimant should reasonably have known and when they should have known it – the fact of loss, that the defendant caused the loss and that legal action was appropriate – have rarely been decided on a class-wide basis. Individual assessments are needed because, as the case law makes clear, the individual claimant’s personal circumstances and knowledge will always be relevant to the reasonable discoverability inquiry.

[56] The Defendant seeks to distinguish these cases by asserting that the courts were applying limitation statutes that altered the common law rule of discoverability. The common law rule would be applicable only in New Brunswick, Newfoundland and Labrador and under the federal limitations regime. In particular, statutes such as Ontario’s Limitations Act, 2002, SO 2002, c 24, sch B, define discoverability as including a requirement that the plaintiff knew or ought to have known that a legal proceeding would be an “appropriate means” to seek remedy. According to the Defendant, the “appropriate means” test is not part of the common law and cases that applied it, such as Novak v Bond, [1999] 1 SCR 808, must be distinguished.

[57] While one must obviously pay attention to the differences between various limitations statutes, this submission does not assist the Defendant. Even assuming that the “appropriate means” test does not form part of the common law rule, it does not follow that the common law rule excludes any consideration of individual circumstances when assessing whether a claimant ought to have known that injury has occurred or what its cause was. Again, the Defendant cited no authority for such a radical proposition.

[58] Moreover, the distinction between the common law rule and its statutory modification is not as clear-cut as the Defendant asserts. In Newfoundland and Labrador, where the Defendant says the common law rule applies, subsection 14(1) of the Limitations Act, SNL 1995, c L-16.1, provides that “the limitation period fixed by this Act does not begin to run against a person until the person knows or, considering all circumstances of the matter, ought to know that the person has a cause of action.” If this is to be considered a codification of the common law, it supports the Plaintiffs’ position that an assessment of objective discoverability must take individual circumstances into account.

[59] The Defendant also asserts that considering any individual component in the objective analysis would erase any distinction between the objective and subjective components of discoverability. This is incorrect. In fact, like the Supreme Court did in Grant Thornton, it may be more appropriate to refer to “actual” and “constructive” knowledge, instead of “subjective” and “objective.” Considering individual circumstances when assessing whether a person ought to have known something is an entirely different question from asking whether the person actually knew. Thus, the distinction between the two prongs of the discoverability test does not collapse.

(2) The Class is Heterogeneous

[60] Even though individual circumstances must be factored into an assessment of objective discoverability, the issue may nevertheless be decided on a class-wide basis if there is evidence that the circumstances of all the members of the class are substantially similar. For example, in Spina, a claim was brought on behalf of a class of franchisees of a drug store chain in respect of the treatment of certain payments pursuant to the franchise agreement. The agreements were the same for every franchisee. The Court found that the controversy about the payments in question was “notorious” in this particular industry and that all franchisees would have realized that the payments were not made when revenues were reconciled at the end of each year. These standardized circumstances allowed the Court to find that all franchisees ought to have discovered their claim as soon as the payments in question became due and not at a later date.

[61] In this case, contrary to Spina, there is no evidence of substantial similarity in the circumstances relevant to the assessment of objective discoverability. Yet, the defendant who brings a motion for summary judgment bears the burden of adducing such evidence: Lameman at paragraph 11. Rather, there is likely to be substantial variation between class members with respect to the time the injury manifested itself (which is a component of the cause of action) and the circumstances that would lead the Court to find that the class members ought to have discovered their cause of action.

[62] In reaching these findings, I rely on the expert report of Dr. Raven Sinclair. I have taken note of the Defendant’s concerns regarding the reliability of certain portions of Dr. Sinclair’s evidence, and I share some of those concerns. Nevertheless, I am confident that the portions of her evidence described below are reliable, especially in light of her long-term and intensive engagement with a large number of Sixties’ Scoop survivors. Moreover, these findings align with common sense.

[63] First, class members may have learned about their adoption at different stages of their lives. Not all adoptive parents were open about class members’ Indigenous identity. Rather, the prevailing attitude favoured the raising of class members as if they were members of the non‑Indigenous society. Both Ms. Varley and Ms. Lukowich stated that their adoptive families did not talk about their Indigenous heritage when they were growing up. In addition, class members may or may not have physical characteristics that would prompt outsiders to recognize them as Indigenous. For example, Ms. Varley’s “apparent race” was described as “white” and “Métis” in two separate medical reports prepared before her adoption, and she reported that her adoptive mother had chosen her among the children available for adoption because she had “the lightest coloured skin [of the AIM children] and would blend with the family”.

[64] Second, there may be significant barriers preventing some class members from accepting that they are Indigenous and attempting to reconnect with their birth families and communities. Again, many class members were raised and socialized as being non-Indigenous and they may value such an identity. Some may have even internalized anti-Indigenous attitudes. In this context, discovering one’s cause of action is synonym with assuming and accepting an identity different from that in which they were raised. I simply cannot assume that each class member’s psychological journey unfolded in a similar manner.

[65] Third, it may be difficult to locate and obtain adoption records and evidence of Indigenous identity or descent, especially where the person is not a status Indian. Ms. Lukowich declared that when she first attempted to retrieve her adoption record, child welfare workers discouraged her from doing so. This difficulty is compounded by the fact that during the Sixties’ Scoop, children were often placed for adoption in another province and sometimes in a different country.

[66] Because of these factors, Dr. Sinclair states that she observed that “the ages at which survivors begin to search for their First Nation or Métis identity, family, community, and culture, ranges from 6–60 years old.” It is relevant to recall that the youngest member of the class would have been 24 years old in 2015.

[67] In my view, all these factors would be highly relevant to an assessment of whether a class member ought to have discovered facts leading to a plausible inference of liability on the part of the federal government by May 2015. These factors are peculiar to each class member. They prevent a class-wide finding regarding objective discoverability.

(3) Earlier Class Actions do not Prove Objective Discoverability

[68] In spite of the wide diversity of the circumstances of class members, the Defendant relies on the bringing of a class action on behalf of Sixties’ Scoop survivors in 2011 to argue that all members of the class ought to have discovered their cause of action by then. The logical fallacy of that proposition is obvious: the fact that one person knew does not entail that all class members ought to have known. Actual and constructive knowledge are separate issues. The former is a factual judgment, while the latter is a normative judgment. The fact that one class member actually found out about their claim does not dispense the Court from assessing whether all class members ought to have done so and does not negate the disparity of the class members’ circumstances. A similar assertion was roundly rejected in Fanshawe College v AU Optronics, 2015 ONSC 2046 at paragraph 92, aff’d 2020 ONCA 621.

[69] The Defendant cites Fehr for the proposition that class counsel’s knowledge can be imputed to class members. It would follow that members of the class in this action, who were also members of the class in the action brought in 2011, would be deemed to have knowledge of their cause of action based on knowledge of counsel who had conduct of the 2011 action. Fehr, however, can be distinguished. It involved the attempt to certify a new cause of action in an existing class action that had been certified for a long time. The facts constituting the new cause of action were common to all class members and class counsel discovered them more than two years before bringing the motion to amend. At first instance, the matter proceeded on the assumption that all class members were bound by class counsel’s knowledge. On appeal, the plaintiffs argued that this proposition was legally unsound. The Ontario Court of Appeal dismissed this argument, mainly because it had not been raised at first instance. Thus, the case is not authority for the proposition that class counsel’s knowledge can be imputed to all class members, even for the purposes of different actions. In any event, Fehr appears to be similar to Spina to the extent that the facts underpinning the cause of action were the same for all class members. As I explained above, this matter is fundamentally different.

[70] For the same reasons, the fact that gatherings of Sixties’ Scoop survivors took place as early as 2007 cannot be held against class members who did not participate. More precisely, the evidence shows that two such gatherings were held in 2014 in Ottawa and Winnipeg and were attended by dozens of persons. However, Plaintiffs’ counsel suggests that the class in the present action could be as large as 20,000 persons. Again, the fact that some class members discovered their claim before 2015 does not entail that all class members reasonably ought to have done so. Class members who have lost connection with their Indigenous community or are not even aware of their Indigenous identity can hardly be expected to attend such gatherings, precisely because they have not yet discovered their cause of action.

(4) Media Reports do not Prove Objective Discoverability

[71] The Defendant also relies on a series of media reports in the period 2011–2015 to show that a reasonable class member would have discovered their claim before May 2015. These reports were published in local newspapers such as the Prince George Citizen, the Edmonton Journal, the Moose Jaw Time Herald and the Winnipeg Free Press, and in newspapers serving the Indigenous community, such as the Windspeaker. Some were distributed by the Canadian Press, and one was found in a local section of CBC News’s website. Generally speaking, they state that Sixties’ Scoop survivors have launched class actions or report on certain steps taken in these actions. They also contain background information about the Sixties’ Scoop. In one case, reference is made to an event held by a group of Sixties’ Scoop survivors in Toronto.

[72] There is little evidence to show the reach of these newspaper reports. I cannot presume that all members of the class have read them. About fifteen press clippings constitute a very tiny portion of the universe of Canadian news over a period of four years. Moreover, the reports are all in English and there is no evidence that they were translated and published in French. Thus, there is no evidence that French-speaking class members would have learned about the Sixties’ Scoop from the media. More generally, there is no evidence that by 2015, the Sixties’ Scoop was sufficiently well-known in the general Canadian public to justify a finding that everyone, including class members, was constructively aware of it.

[73] The Defendant did not cite any case where media attention of such a relatively low intensity led to a finding of objective discoverability. In Smith, for example, much more intensive media coverage, focused on a single, small community, together with an array of outreach measures, was still found insufficient to ground a finding that all members of the group ought to have known about their claim.

D. Summary on Limitations

[74] For the foregoing reasons, the Defendant has failed to make out its defence that all members of the claim ought to have discovered their cause of action before May 2015, that is, more than six years before the certification of the action. Accordingly, the Defendant’s motion for summary judgment will be dismissed.

IV. Duty of Care and Fiduciary Duty

[75] We can now turn to the Plaintiffs’ motion for summary judgment. The Plaintiffs are seeking a declaration that the federal government owed a duty of care or a fiduciary duty to class members. For the reasons that follow, I am granting their motion in part only.

[76] With respect to the entire class, the Plaintiffs’ claim fails because the federal government’s alleged omissions resulted from a core policy decision that is immune from liability in tort. In any event, there is no duty of care because the apprehension, placement and adoption of Métis and non-status Indian children took place without any direct or indirect intervention by the federal government. There was no fiduciary duty, largely because the federal government did not assume any form of discretionary control over Indigenous culture and identity with respect to Métis and non-status Indian children.

[77] Nevertheless, the federal government had a duty of care toward children who were placed or adopted through Saskatchewan’s AIM program. In the particular circumstances of the case, federal funding of the program created the requisite proximity and the harm to the children was foreseeable.

A. Preliminary Issues

[78] Before analyzing whether the Defendant owed a duty of care or a fiduciary duty to class members, it is necessary to clarify certain aspects of the broader legal framework governing the claim. The Plaintiffs have framed their claim in private law terms. The extent to which notions of reconciliatory justice developed in the context of public law can be applied to private law claims must first be ascertained. The fact that claims are framed in terms of private law also means that the Court must base its decision on the private law applicable in a particular province or territory, namely, civil law in Quebec and common law elsewhere in Canada. Whether the record is sufficient to resolve the issues by way of a summary judgment must also be ascertained.

(1) Reconciliatory and Corrective Justice

[79] This case puts into contrast two dimensions of justice, which can be called corrective and reconciliatory justice: Quebec (Attorney General) v Pekuakamiulnuatsh Takuhikan, 2024 SCC 39 at paragraph 148 [Pekuakamiulnuatsh Takuhikan]. Corrective justice is mainly embodied in doctrines of private law. Its aim is to correct the breach of a pre-existing legal obligation by putting the parties in the position in which they would have found themselves had there not been a breach. Given its focus on a breach, it is mainly fault-based or, in other words, focused on wrongdoing. Reconciliatory justice, in contrast, aims at “repairing and maintaining the special relationship with the Indigenous peoples” (ibid). Its scope is broader and may encompass historic and collective injustices that cannot be fully captured by the tools of corrective justice. The development of doctrines of public law regarding Indigenous peoples has been strongly influenced by concerns for reconciliatory justice: see, for example, Haida Nation v British Columbia (Minister of Forests), 2004 SCC 73 at paragraph 27, [2004] 3 SCR 511.

[80] Corrective justice is the heartland of the courts’ mission. Courts can also contribute to reconciliatory justice in appropriate cases. For example, the honour of the Crown is a concept frequently used for this purpose. Nevertheless, other legal tools, such as legislation or negotiation, may be better suited to the achievement of reconciliatory justice: R v Desautel, 2021 SCC 17 at paragraph 87, [2021] 1 SCR 533 [Desautel]; Shot Both Sides at paragraph 71. In these cases, reconciliatory justice is not hampered by doctrines associated with corrective justice, such as limitation periods or the immunity associated with policy decisions.

[81] For instance, it may well be that the federal government was inspired more by reconciliatory than corrective justice when it settled certain class actions brought by Indigenous claimants, as my colleague Justice Peter Pamel (now of the Federal Court of Appeal) recently noted in Percival v Canada, 2024 FC 2098 at paragraph 40:

. . . by 2018, . . . Canada was looking towards reconciliation, and where the research confirmed evidence of hardship occasioned by the misguided Indigenous children educational policies of the past, it was more likely than not that Canada would be willing to sit, listen and discuss a way forward, without aggressive posturing or a vigorous defence of the claims. To a great extent, by 2018, Canada had moved beyond being an adversary in litigation of this type to, in essence, being a willing partner in seeking a resolution and reconciliation for the harms that had been committed.

[82] As this excerpt suggests, reconciliation may require harm-based compensation schemes that private law doctrines based on fault and corrective justice cannot always provide. This is not to suggest that private law is impervious to reconciliatory justice. Private law doctrines are flexible enough to adapt to a wide variety of circumstances, including those of Indigenous peoples. However, they cannot be stretched to a point where they lose their essential character.

[83] In this case, as I will explain below, granting judgment in favour of the entire class would require disregarding the well-established immunity pertaining to policy decisions and jettisoning the requirement of proximity, which lies at the heart of tort law. This would effectively entail a finding of liability without fault, which is inimical to the principles of tort law. Hence, if reconciliation requires compensation for all class members, this will need to be addressed outside the judicial forum. Nevertheless, as explained below, finding a duty of care in favour of the class members who were placed and adopted through the AIM program is an exercise of reconciliatory justice that properly belongs to the judicial role.

(2) Common Law and Civil Law

[84] The class comprises members living everywhere in Canada. Moreover, some class members may have been placed in a province other than their province of birth. It is therefore necessary to map the respective domains of the common law and the civil law in the determination of the parties’ rights. Because this is a claim for damages against the federal Crown, section 3 of the Crown Liability and Proceedings Act applies:

3 The Crown is liable for the damages for which, if it were a person, it would be liable	3 En matière de responsabilité, l’État est assimilé à une personne pour :
(a) in the Province of Quebec, in respect of	a) dans la province de Québec :
(i) the damage caused by the fault of a servant of the Crown . . .	(i) le dommage causé par la faute de ses préposés […]
(b) in any other province, in respect of	b) dans les autres provinces :
(i) a tort committed by a servant of the Crown . . .	(i) les délits civils commis par ses préposés […].

[85] Section 3 is consistent with the broader principle established by the Quebec Act, 1774, to the effect that matters of “property and civil rights” in Quebec are governed by the civil law, not the common law. Therefore, unless a public law rule explicitly provides otherwise, the private law relationships of governments and public institutions are governed by the private law of the province concerned: Prud’homme v Prud’homme, 2002 SCC 85 at paragraph 46, [2002] 4 SCR 663. Where the application of federal legislation is at stake, this principle is embodied in section 8.1 of the Interpretation Act, RSC 1985, c I-21.

[86] Indeed, in their statement of claim, the Plaintiffs asserted that where the acts of Canada’s servants took place in Quebec, they gave rise to extracontractual liability pursuant to article 1457 of the Civil Code of Québec, instead of liability in tort at common law. Nevertheless, in both their written and oral submissions, the Plaintiffs adopted an entirely different position and argued that the case is wholly governed by “federal common law” and that the civil law plays no role, even where the cause of action arose in Quebec.

[87] I am unable to agree with the Plaintiffs. Except perhaps in admiralty matters, there is no such thing as a “federal common law,” if one means by that term a freestanding and comprehensive body of private law that displaces provincial law in certain areas; see, for example, Quebec North Shore Paper v CP Ltd, [1977] 2 SCR 1054 at 1065; Desgagnés Transport Inc v Wärtsilä Canada Inc, 2019 SCC 58 at paragraph 47, [2019] 4 SCR 228; H Patrick Glenn, “The Common Law in Canada” (1995) 74 Can Bar Rev 261, at 279–280. The Plaintiffs have framed their claims in private law terms, namely the tort of negligence and fiduciary duty. These claims are captured by section 3 of the Crown Liability and Proceedings Act and even if they were not, the more general principle that, subject to statutory exceptions, the federal Crown’s private law relationships are governed by the law of the province concerned requires the application of Quebec civil law where the cause of action arises in that province.

[88] Relying on Roberts v Canada, [1989] 1 SCR 322, the Plaintiffs argue that the present matter is governed by “federal common law” because it is concerned with the federal government’s relationship with Indigenous peoples. However, that case does not assist the Plaintiffs. The claim in that case pertained to the possession of reserve land and the Crown’s fiduciary duty towards the two First Nations contending for such possession. The jurisdiction of the Federal Court to hear the matter depended on whether the claim was based on a “law of Canada” within the meaning of section 101 of the Constitution Act, 1867. The Supreme Court found that the claim was based in part on the common law of aboriginal title. It further decided that this body of common law was federal, apparently because it was incorporated by reference in the Indian Act, a piece of federal legislation.

[89] There is little reason to extend Roberts beyond what it actually decides. The law of aboriginal title is not the basis for the cause of action put forward in this action. Aboriginal title pertains to public law (or “Imperial law”), not private law, which explains its uniform application throughout Canada: Brian Slattery, “Understanding Aboriginal Rights” (1987) 66 Can Bar Rev 727 at 737–739. There is, however, no basis for a more general proposition that private law claims brought by Indigenous peoples in Quebec are governed by “federal” common law. In Pekuakamiulnuatsh Takuhikan, for example, the Supreme Court applied the rules of the civil law to a contract between a First Nation and the federal and provincial governments. Likewise, provincial limitation statutes apply to claims brought by Indigenous peoples and provincial legislation regarding contributory negligence applies to claims related to residential schools: Blackwater v Plint, 2005 SCC 58 at paragraphs 66–67.

[90] Nor does the “national” dimension of the present class action have any bearing on the applicable law. A class action is a procedural vehicle that does not affect the law applicable to the substance of each member’s claim: Bisaillon v Concordia University, 2006 SCC 19 at paragraph 17, [2006] 1 SCR 666. Where the class comprises members residing in Quebec and other provinces or territories, or where the claim encompasses causes of action arising in Quebec and elsewhere, the court must ascertain which claims are governed by Quebec law and decide the case accordingly.

[91] The Plaintiffs’ stance leaves the Court in an unfortunate position. They declined to make alternative submissions in the event Quebec law were applicable and the Defendant did nothing to assist. I cannot simply assume that the result under Quebec law would be the same as under the common law, for this would amount to applying the common law in Quebec; see, in this regard, Thompson v Canada, 2025 FC 476 at paragraphs 194–195.

[92] The parties’ failure to make meaningful submissions regarding civil law means that I am not in a position to decide the motion with respect to causes of action arising in Quebec. I must then dismiss the motion for summary judgment with respect to these causes of action.

(3) Availability of Summary Judgment

[93] The Defendant objects to the Plaintiffs’ choice to ask the Court to decide only the issue of the duty of care in this motion for summary judgment. Although it is fairly common to bifurcate liability and damages, the Defendant argues that bifurcating duty of care and standard of care is unprecedented and unfair because it calls upon the Court to determine a crucial issue in a factual vacuum.

[94] In my view, there is nothing improper in asking the Court to rule on the duty of care only. Rule 213 of the Federal Courts Rules, SOR/98-106, provides that a motion for summary judgment may pertain to “all or some of the issues raised in the pleadings.” There is no indication that the availability of summary judgment on “some of the issues” is limited to the bifurcation between liability and damages. The Defendant’s submission that “there is no established precedent for the novel bifurcation of the existence of a duty of care from the question of whether there has been a breach of the standard of care” overlooks the Supreme Court’s recent decision in 1688782 Ontario Inc v Maple Leaf Foods Inc, 2020 SCC 35, [2020] 3 SCR 504, arising from a motion for summary judgment dealing with duty of care only. Seeking summary judgment with respect to the duty of care does not give rise to unfairness. The evidence relevant to the duty of care issue pertains to the relationship between the parties. The evidence pertaining to the standard of care would cover different issues and is not needed to determine the existence of a duty of care.

[95] More generally, I am satisfied that the evidence enables me “to reach a fair and just determination on the merits:” Hryniak v Mauldin, 2014 SCC 7 at paragraph 49, [2014] 1 SCR 87; Saskatchewan (Attorney General) v Witchekan Lake First Nation, 2023 FCA 105 at paragraphs 30–40. With respect to duty of care, the case is largely based on archival and historical documents. In all likelihood, government officials involved in the relevant discussions and decisions have all passed away. Credibility is not in issue. Beyond the Defendant’s general objection to the bifurcation between duty of care and standard of care, the parties have not suggested that additional evidence would be available at trial or that the matter did not lend itself to summary judgment. Hence, proceeding by way of summary judgment is appropriate.

B. Duty of Care With Respect to the Entire Class

[96] The Plaintiffs first assert a cause of action based on the tort of negligence. In this kind of claim, the plaintiff must show that the defendant had a duty of care towards them; that the defendant breached that duty by failing to act according to the relevant standard of care; that the plaintiff sustained damage; and that the damage was caused by the breach: Mustapha v Culligan of Canada Ltd, 2008 SCC 27 at paragraph 3, [2008] 2 SCR 114 [Mustapha]. In their motion for summary judgment, the Plaintiffs are only asking the Court to rule on the first element of this test, the duty of care.

[97] Canadian common law courts assess the duty of care according to what is known as the Anns/Cooper framework. This framework comprises two stages, described in cases such as Cooper v Hobart, 2001 SCC 79, [2001] 3 SCR 537 [Cooper]; Deloitte & Touche v Livent Inc (Receiver of), 2017 SCC 63, [2017] 2 SCR 855 [Livent]. At the first stage, the plaintiff must prove that the defendant had a “prima facie duty of care” by showing that there was sufficient proximity between the plaintiff and defendant and that the defendant could reasonably foresee that their conduct could cause harm to the plaintiff. At the second stage, the defendant may attempt to show that the prima facie duty of care is negated by policy considerations, including an immunity afforded to “core policy decisions:” Nelson (City) v Marchi, 2021 SCC 41, [2021] 3 SCR 55 [Marchi].

[98] In this section, I will analyze whether the federal government had a duty of care towards the entire class. I conclude that it did not, because the omissions that allegedly caused harm to the members of the class resulted from core policy decisions and there is no proximity between the federal government and the entire class. In the next section, however, I will show that it had a duty of care towards children placed in foster care or adopted through Saskatchewan’s AIM program.

(1) Immunity for Core Policy Decisions

[99] Immunity for policy decisions is usually considered at stage two of the Anns/Cooper test, after proximity and foreseeability have been established. In this case, however, it is more useful to address the issue of immunity first, as it provides a complete answer to the claim.

[100] The federal government did not have a duty of care towards the entire class because the omissions that the Plaintiffs plead were the result of a policy decision that benefits from an immunity in tort.

(a) Nature of the Decision

[101] In their statement of claim, the Plaintiffs allege that class members suffered harm because of certain omissions of the federal government. These omissions include the failure to ensure that provincial child welfare systems were adequate and did not deprive class members of their culture and identity, and the failure to take measures to mitigate the impact of apprehension, placement and adoption on class members. These omissions, however, were the product of a more general policy decision, namely, the decision not to assume responsibility for the provision of services to the Métis and non-status Indian population, where these services were already offered by the provinces to the general population.

[102] While this policy was often not stated explicitly, it flows, a contrario, from the policy to fund services only for status Indians living on reserves, as reflected in the numerous funding agreements filed in evidence by both parties. Nonetheless, and leaving aside the special case of the Inuit, it is clear that the federal government did not accept responsibility for the provision of services to Indigenous persons who did not have Indian status. It assumed that these persons would receive services from the provinces, usually on the same basis as other residents, as the provinces were generally not inclined to make distinctions based on Indigenous identity or ancestry in the provision of services. Even though the federal government had internal legal opinions to the effect that Parliament could legislate for Métis and non-status Indians pursuant to section 91(24) of the Constitution Act, 1867, it considered that the provinces were chiefly responsible for the provision of services to these groups. Indeed, this was consistent with the policy of integration pursued by the federal government and embodied in the 1951 reform of the Indian Act and the 1969 White Paper.

[103] The evidentiary record suggests that this policy was in force throughout the entire claim period, from 1951–1991. In a memorandum to Cabinet dated August 10, 1959, the Minister of Citizenship and Immigration noted that it had been the objective of her department to secure for status Indians the same welfare services enjoyed by other citizens and that it was impractical for the federal government to duplicate provincial services. She then stated the extent to which the federal government was prepared to assume liability:

3. Indians who have established residence in non-Indian communities are liable for the same taxes as non-Indians and, in the view of the Department, should have access to all community and provincial services on the same basis as non-Indians. There seems to be no case for financial participation by the Federal Government in regard to welfare services on behalf of Indians in such circumstances.

4. In regard, however, to Indians on reserves and in their own communities, the Department of Citizenship and Immigration has assumed responsibility for essential welfare services in the past. Extension of provincial services to these areas, therefore, entails financial agreements with the provinces in regard to the cost of benefits and the costs of administration.

[104] This policy was also described in a 1980 discussion paper by the Department of Indian Affairs and Northern Development [DIAND] entitled Natives and the Constitution, which noted that the federal government “has further chosen, as a matter of policy, to limit its responsibility for the provision of direct services to status Indians, under the Indian Act, essentially to services provided on reserve.” When describing the assumptions that could inform a future federal position, Natives and the Constitution made it clear that “Métis and non-status Indians would continue to fall in large but [sic] under the same program arrangement as non-native citizens of the provinces”.

[105] With respect to child welfare, the policy is best described in a program circular of the DIAND dated May 1, 1982:

3.2 Provincial or Territorial Governments have a legal responsibility to provide care and protection to dependent and neglected children residing within their geographic boundaries.

3.3 DIAND has accepted financial responsibility and has authority to reimburse, as per agreement, Provincial and Territorial Governments and accredited child care agencies, for the cost of child welfare services to Indian children and parents residing on reserve or Crown land and unorganized territories.

[Emphasis in original]

[106] In this circular, “dependent and neglected children residing within [a province’s] geographic boundaries” includes Métis and non-status Indian children.

[107] By and large, the federal government acted consistently with this policy in relation to child welfare services during the period covered by this action. The federal-provincial agreements providing for the reimbursement of the costs incurred by the provinces in applying their child welfare legislation pertained to status Indians living on reserves or in communities akin to reserves. The Plaintiffs pointed to agreements with Alberta and Nova Scotia extending coverage to children without status living on reserves in certain circumstances, as well as agreements with Newfoundland and Labrador that provided federal funding before Indigenous peoples of the province were brought under the Indian Act or formally recognized as Inuit. These situations, however, amounted to an adaptation of the policy to exceptional circumstances. They cannot be interpreted, as the Plaintiffs contend, as a general recognition of federal responsibility for the provision of services to Métis and non-status Indians. Likewise, the policy described above is not inconsistent with the provision of federal services to Métis and non-status Indians in areas in which the federal government directly provides services to the general population, such as housing, workforce training or regional and local economic development.

(b) Does the Immunity for Core Policy Decisions Apply?

[108] Since at least Kamloops (City of) v Nielsen, [1984] 2 SCR 2, Canadian law has recognized that governments and public authorities are immune from liability in tort with respect to “core policy decisions.” The rationale is that “each branch of government has a core institutional role and competency that must be protected from interference by the other branches”: Marchi at paragraph 3. As the Supreme Court further explains in the latter case, at paragraphs 44–45:

Core policy decisions of the legislative and executive branches involve weighing competing economic, social, and political factors and conducting contextualized analyses of information. These decisions are not based only on objective considerations but require value judgments — reasonable people can and do legitimately disagree . . .

Relatedly, the adversarial process and the rules of civil litigation are not conducive to the kind of polycentric decision-making done through the democratic process . . .

[109] Still in Marchi, at paragraphs 61–65, the Court explained that there are four factors to be examined to determine whether a government decision is a policy decision. I will structure my analysis according to these four factors.

[110] The first factor is the level and responsibilities of the decision-maker. Here, it appears that the policy to assume responsibility only in respect of status Indians was made at the highest levels of government. For example, a Treasury Board memorandum dated January 12, 1972 recommended that the Governor in Council approve the entry into agreements with provinces with respect to “the extension of provincial welfare programs to Indians” and there is a reference to earlier decisions made by Cabinet in this respect. The federal-provincial agreements entered into evidence are typically signed by federal and provincial ministers, which tends to indicate that they relate to important policy matters.

[111] The second factor is the process by which the decision was made. There is no evidence of when the policy was first established. The January 12, 1972 memorandum, however, shows that the policy and potential changes to it were the product of discussions between several government departments, in a process typical of policy development. Likewise, the discussion paper Natives and the Constitution shows that any changes to the policy would involve consultations with provinces and Indigenous organizations. Moreover, to borrow language from the Supreme Court in Marchi, at paragraph 63, the policy “was intended to have broad application and be prospective in nature;” it was not “a reaction of an employee or groups of employees to a particular event.” This tends to show that it is shielded from liability.

[112] The third factor is the nature and extent of budgetary considerations. It is obvious that the choice of the category of persons over whom the federal government accepted responsibility had significant budgetary impacts. For present purposes, it is not necessary to quantify these impacts. It is nevertheless obvious from the documentary evidence that the extension of federal funding to Métis and non-status Indians would produce a significant increase in the population under federal responsibility. It can easily be surmised that the impact would be in the tens of millions of dollars yearly, during the period covered by the action, in respect of child welfare services only. Impacts of that magnitude are typical of a core policy decision.

[113] The fourth factor is “the extent to which the decision was based on objective criteria:” Marchi at paragraph 65. Here, the decision was not based on any objective criteria, which tends to show that it is a core policy decision. While the usual assumption in Canadian public administration is that financial responsibility is congruent with legislative jurisdiction, there is no principled formula for allocating financial responsibility in double aspect areas, that is, areas in which both Parliament and the provinces may legislate. In the case of child welfare for Indigenous peoples, the division of responsibility was the result of unilateral assertions and negotiation between the federal government and the provinces, not the application of any objective criteria. Natives and the Constitution and other policy documents of the era show that this was a polycentric issue involving a wide array of considerations, which is typical of core policy decisions immune from liability in tort.

[114] Hence, the federal government’s alleged omissions that form the basis of the asserted cause of action were the direct result of a core policy decision with respect to the definition of categories of Indigenous persons to whom the federal government was prepared to provide services. The policy and the omissions that are its direct consequence are therefore immune from liability in tort.

(c) Was the Decision Irrational or Made in Bad Faith?

[115] Relying on R v Imperial Tobacco Canada Ltd, 2011 SCC 42 at paragraph 90, [2011] 3 SCR 45 [Imperial Tobacco], the Plaintiffs nevertheless argue that the decision not to assume responsibility for Métis and non-status Indians should not benefit from a policy immunity because it was irrational or made in bad faith. According to the Plaintiffs, the decision would be irrational because it relied on identity categories that were themselves known to be irrational or arbitrary and it would be in bad faith because its purpose was mainly to save money.

[116] These considerations do not take the decision out of the sphere of core policy decisions immune from liability. They amount to an invitation for the Court to rule as to the merits of the policy in question. Yet, immunity is granted precisely to ensure that the merits of policy decisions remain a matter for the executive and legislative branches of government.

[117] In this context, “bad faith” and “irrationality” are very high bars. The fact that a decision would be vulnerable to judicial review is not enough: Entreprises Sibeca Inc v Frelighsburg (Municipality), 2004 SCC 61 at paragraph 23, [2004] 3 SCR 304 [Sibeca]. Liability would ensue only where an act is “inexplicable and incomprehensible, to the point that it can be regarded as an actual abuse of power, having regard to the purposes for which it is meant to be exercised”: Finney v Barreau du Québec, 2004 SCC 36 at paragraph 39, [2004] 2 SCR 17. In other words, acts would fall outside the scope of immunity if they “are so markedly inconsistent with the relevant legislative context that a court cannot reasonably conclude that they were performed in good faith”: Sibeca at paragraph 26.

[118] It is no secret that many of the assumptions behind the policies of the period 1951–1991 have now changed. The rules of Indian status have undergone several rounds of amendments to root out various forms of discrimination. With respect to child welfare, cultural continuity is now considered a directing principle, setting aside the philosophy of integration epitomized by Racine and similar cases. In the wake of Daniels, the federal government has been more willing to take initiatives with respect to the Métis, and Bill C-92 treats all Indigenous peoples in the same manner. Indeed, the point made in Natives and the Constitution, that Métis and non-status Indians are subject to many of the disadvantages suffered by status Indians, is increasingly accepted.

[119] The fact that the policies of the past have changed or that they are now perceived as wrongful does not, however, mean that they were irrational or in bad faith, so that they would no longer be immune from liability. Consider Imperial Tobacco, which dealt with the federal government’s promotion of low-tar cigarettes for those who did not want to stop smoking. This policy of the late 1960s was later found to be misguided and harmful, yet the Supreme Court concluded that it benefitted from the immunity afforded to core policy decisions. Likewise, there is no evidence of bad faith or irrationality in the federal government’s policy of funding child welfare services for status Indians children living on reserves only and letting the provinces fund services for other Indigenous children. Again, many aspects of this policy have now changed, but irrationality and bad faith are very high bars and the evidence does not justify such a finding in the present case. I hasten to add that this is not a situation where a benefit is conferred on one category of Indigenous persons and withheld from another. Contrary to the Caring Society case, there is no allegation of discrimination. Moreover, the Plaintiffs are not raising any constitutional challenge to the policy.

[120] The Plaintiffs argue that the federal government was in bad faith because it merely sought to save money at the expense of Métis and non-status Indians. Yet, as we have seen above, financial considerations are one of the hallmarks of core policy decisions and cannot logically suffice to take these decisions out of the sphere of immunity. Here, the federal government’s policy did not deprive Métis and non-status Indians of child welfare services. Rather, it adopted a policy that assumed that these services would be offered by the provinces.

[121] To summarize, the omissions on which the Plaintiffs ground their claim are the product of a core policy decision that was neither irrational nor made in bad faith. As a result, the federal government can claim an immunity from liability that negates any duty of care.

(2) Proximity

[122] The above is sufficient to dispose of the duty of care issue with respect to the entire class. Nevertheless, as the parties made extensive submissions regarding proximity, I will also address this issue.

[123] The concept of proximity is “used to describe the type of relationship in which a duty of care to guard against foreseeable negligence may be imposed”: Cooper at paragraph 32. The case law establishes several categories of relationships where a duty of care presumptively arises, for example a municipality’s duty towards the users of a public road. Where the situation does not fit within one of the established categories, however, the court must decide whether a prima facie duty of care exists based on a review of all relevant circumstances, which may include “expectations, representations, reliance, and the property or other interests involved”: Cooper at paragraph 34. Where the defendant is a public body, the relevant statutory framework must also be considered: Syl Apps Secure Treatment Centre v BD, 2007 SCC 38 at paragraphs 27–29, [2007] 3 SCR 83; Imperial Tobacco at paragraphs 43–45. The aim of the inquiry is to determine whether the plaintiff is “closely and directly affected” by the defendant’s act so as “to make it just and reasonable to impose an obligation on one party to take reasonable care not to injure the other”: Cooper at paragraph 32, quoting from Donoghue v Stevenson, [1932] AC 562 (HL) at 580–581 [Donoghue]; Imperial Tobacco at paragraph 41.

[124] In this section, I will first examine the Plaintiffs’ submission that this case falls into a category established by Brown. I will then review the main factors the parties put forward for or against proximity.

(a) An Established Category?

[125] The Plaintiffs first argue that their claim fits within an established category because a duty of care was found to exist in similar circumstances in Brown. That case, however, can be distinguished from the present one, because the factors on which the Court relied to find proximity are not present here: Livent at paragraph 28.

[126] The class members in Brown were status Indians residing on reserve. By and large, the Ontario government did not apply its child welfare legislation to these persons until the federal government agreed to provide funding. The relationship between the federal government and class members in Brown was fundamentally different from what it was in the present case—the provinces applied their child welfare legislation to non-status and Métis children without asking for any federal funding. Moreover, the duty of care in Brown was mainly based on the breach of a term of the federal-provincial agreement requiring consultation with the First Nations concerned; no such thing happened in this case.

[127] I acknowledge that Justice Belobaba relied on the historical relationship between Indigenous peoples and the federal government as an alternative ground for finding a duty of care. His relatively short reasons in this regard must be read in light of the circumstances of the case. The class members were status Indian children living on reserve. Ontario was unwilling to apply its child welfare laws to these children unless the federal government provided funding. Therefore, federal funding was instrumental in bringing about the Sixties’ Scoop in respect of these children. Thus, Justice Belobaba must have grounded his finding in the effects of targeted federal funding on the transmission of Indigenous identity and culture in the context of the historical relationship between the federal Crown and Indigenous peoples. Save in respect of the AIM program, discussed below, this combination of factors is not present in this case. In contrast, reading Justice Belobaba’s comments out of context would give rise to concerns regarding indeterminate liability: Cooper at paragraph 37; Livent at paragraphs 42–45.

[128] Thus, the present case does not fit within a “category” established by Brown and must be analyzed as a novel claim.

(b) Importance of Interest Involved

[129] The nature and importance of the plaintiff’s interest affected by the defendant’s actions is a relevant factor in the proximity analysis: Cooper at paragraph 34. There is no doubt that the interests involved in this action, namely, Indigenous identity and culture, are significant. They were protected by law during the period covered by the claim.

[130] Belonging to an Indigenous group is a fundamental component of one’s individual identity. Likewise, connection with one’s Indigenous community will usually bring a sense of psychological security, access to land and its teachings and other benefits flowing from kinship (or wáhkôhtowin). Courts have taken notice of the importance of identity and community connections. For example, the Supreme Court recognized that “relationships within Indian families and reserve communities [are] matters that could be considered absolutely indispensable and essential to their cultural survival”: Canadian Western Bank v Alberta, 2007 SCC 22 at paragraph 61, [2007] 2 SCR 3. The Supreme Court of British Columbia has also held that official recognition of one’s Indigenous identity reinforces “a sense of identity, cultural heritage, and belonging”: McIvor v The Registrar, Indian and Northern Affairs Canada, 2007 BCSC 827 at paragraph 286 (and paragraphs 123–143), aff’d 2009 BCCA 153 at paragraph 70. In short, the recognition of Indigenous identity is linked to psychological integrity, an interest protected by tort law: Mustapha at paragraph 8; Saadati v Moorhead, 2017 SCC 28 at paragraph 23, [2017] 1 SCR 543.

[131] In addition, courts have found that various aspects of Indigenous culture and identity are recognized and affirmed by section 35 of the Constitution Act, 1982. For example, in R v Côté, [1996] 3 SCR 139 at paragraph 56 [Côté], the Supreme Court held that cultural transmission would usually be an incident of aboriginal rights. More recently, courts have begun to recognize that components of Indigenous culture and identity can be generic aboriginal rights recognized by section 35. For example, in Caring Society, at paragraph 106, the Canadian Human Rights Tribunal stated:

. . . the specific Aboriginal interests that stand to be adversely affected in this case are, namely, indigenous cultures and languages and their transmission from one generation to the other. Those interests are also protected by section 35 of the Constitution Act, 1982. The transmission of indigenous languages and cultures is a generic Aboriginal right possessed by all First Nations children and their families.

[132] See also Renvoi à la Cour d’appel du Québec relatif à la Loi concernant les enfants, les jeunes et les familles des Premières Nations, des Inuits et des Métis, 2022 QCCA 185 at paragraphs 468–494 [Quebec Reference]; Eskasoni First Nation v Canada (Attorney General), 2024 FC 1856 at paragraphs 79–81; Fisher River Cree Nation v Canada (Attorney General), 2025 FC 561 at paragraph 75. In Reference re An Act respecting First Nations, Inuit and Métis children, youth and families, 2024 SCC 5 [Bill C-92 Reference], the Supreme Court did not explicitly rule on the scope of the rights recognized by section 35, but it clearly appreciated the importance of the transmission of Indigenous culture and identity and the maintenance of community connections.

[133] Rights related to identity and culture were not created by section 35 of the Constitution Act, 1982. They existed before, even though they were often ignored: Côté at paragraphs 52–54; Desautel at paragraph 34. Thus, Indigenous culture and identity were legally protected interests during the period covered by the claim.

[134] Moreover, aboriginal rights are collective rights that an Indigenous person cannot meaningfully exercise if they are estranged from their community or, a fortiori, if they do not know they are Indigenous. For instance, courts have acknowledged that the Métis have aboriginal harvesting rights recognized and affirmed by section 35 of the Constitution Act, 1982: R v Laviolette, 2005 SKPC 70; R v Belhumeur, 2007 SKPC 114; R v Goodon, 2008 MBPC 59; R v Boyer, 2022 SKCA 62. Children removed from their Métis families and communities are deprived of a meaningful opportunity to exercise such rights.

[135] Thus, the interests of class members who were affected by the Sixties’ Scoop were fundamental and recognized by law. Nonetheless, the fact that these interests were affected is insufficient to ground proximity. The relationship between the Plaintiffs and Defendant must also be considered. I now turn to this issue.

(c) Lack of Direct or Indirect Federal Involvement

[136] The Defendant’s position is largely based on the fact that the federal government had no direct or indirect relationship with class members with respect to their apprehension, placement in a foster family and adoption. Rather, this was done pursuant to provincial legislation applied by provincial employees or organizations deriving their powers from provincial legislation, such as children’s aid societies. Moreover, subject to one exception described below, the federal government did not fund the provinces’ application of their child welfare laws to Indigenous persons other than status Indians residing on reserves, who are already included in the Riddle settlement. I agree that this is a factor that strongly militates against a finding of proximity.

[137] The Plaintiffs try to circumvent this difficulty by highlighting the fact that the statement of claim is based on the omission of the federal government to take measures to protect their identity and culture, which were jeopardized by the application of provincial legislation. A claim based on an omission, so the argument goes, cannot logically be defeated by arguing a lack of direct interactions.

[138] Regardless of the merits of such an argument in other contexts, it cannot apply in the present case where the essence of the claim is that the federal government should have done something to prevent the provinces from negatively impacting Indigenous culture and identity. Unless additional factors are present, there cannot be such a form of “interjurisdictional proximity.” In our constitutional system, the federal and provincial governments are independent of each other: Maritime Bank (Liquidators of) v New Brunswick (Receiver General), [1892] AC 437 (PC); Reference re Secession of Quebec, [1998] 2 SCR 217 at paragraphs 55–60. One order of government is not automatically liable for the other’s actions or omissions. Proximity must be established separately in relation to each one.

[139] In particular, it is difficult to understand how the federal government can become liable for the manner in which provincial governments apply provincial laws, in the absence of any federal legislation or funding. It would run against the principle of federalism for one order of government to have a duty of care that derives purely from the actions or omissions of the other. Such a duty of care would require one order of government to monitor and police the work of the other. This would be inimical to the autonomy of each order of government in our constitutional system. Moreover, short of enacting legislation in areas of double aspect, it is unclear how one order of government may prevent the other from applying its own laws. This is true even where Indigenous peoples are involved. Section 91(24) of the Constitution Act, 1867 does not give the federal government a supervisory role regarding the interaction between provinces and Indigenous peoples: Grassy Narrows First Nation v Ontario (Natural Resources), 2014 SCC 48 at paragraph 30, [2014] 2 SCR 447; George Gordon First Nation v Saskatchewan, 2022 SKCA 41 at paragraphs 160–162.

[140] While the federal government could conceivably become liable in tort for the actions of a province where it undertakes a joint venture with the province or otherwise acts in a manner that establishes proximity with the plaintiff, the evidence in the present case does not support such a finding with respect to the entire class. Rather, subject to the exception analyzed below, the evidence shows that the apprehension and placement of Métis and non-status children during the Sixties’ Scoop was entirely the result of provincial action and did not involve any form of federal action. In this regard, the Ontario Court of Appeal found that there was no proximity between the Ontario government and children apprehended by a Children’s Aid Society, because Ontario was “too far removed from the daily operation of the CAS in child protection matters to give rise to a duty of care”: JB v Ontario (Child and Youth Services), 2020 ONCA 198 at paragraph 53 [JB]. Absent other factors establishing proximity, this holding is applicable to the situation of the entire class in the present case.

[141] Moreover, contrary to the situation of status Indians living on reserve, there is no historical evidence suggesting that the provinces ever refused to apply their child welfare laws to Métis and non-status children. The fact that such children were in fact apprehended and placed in the absence of federal-provincial funding agreements reinforces this conclusion.

[142] The above would normally be dispositive of the issue of proximity. The Plaintiffs, however, submit that proximity can be grounded in either the Crown’s historical relationship with Indigenous peoples or in Parliament’s legislative jurisdiction over them. I now turn to these submissions.

(d) Historical Relationship

[143] The Plaintiffs assert that the federal government has an overarching duty to protect all Indigenous peoples, in particular against loss of culture or identity, irrespective of state-created identity categories. In particular, the federal government would be required to intervene where provincial laws or policies are likely to have a detrimental effect on Indigenous identity and culture. This duty would flow from the historical relationship between the federal government and Indigenous peoples, as evidenced by a series of treaties and other commitments by which the federal government undertook to protect Indigenous peoples’ way of life. It would also flow from the purpose of section 91(24) of the Constitution Act, 1867, which would be the protection of Indigenous peoples against local settlers or provincial governments. According to the Plaintiffs, this duty would arise “as a matter of history, as a matter of constitutional obligation under s. 91(24) of the Constitution Act, 1867, and as a result of the honour of the Crown.”

[144] In my view, while there is obviously a historical relationship between the federal government and Indigenous peoples, it does not automatically give rise to the finding of proximity that is necessary to ground a private law duty of care. If it were otherwise, the federal government’s potential liability would be unbounded. Liability in tort would in effect be transformed into a harm-based form of liability. In reality, the Plaintiffs are asking me to deploy a form of reconciliatory justice that is outside the bounds of private law. Nevertheless, the historical relationship may be relevant to the analysis of other indicia of proximity.

[145] In this regard, attempts to ground private law liability in the relationship alone have been rejected. For example, with respect to fiduciary duties, the Supreme Court repeatedly held that while the relationship between the state and Indigenous peoples is fiduciary in nature, not every aspect of that relationship gives rise to a fiduciary duty enforceable by the courts: Wewaykum Indian Band v Canada, 2002 SCC 79 at paragraph 83, [2002] 4 SCR 245 [Wewaykum]; Manitoba Metis Federation Inc v Canada (Attorney General), 2013 SCC 14 at paragraph 48, [2013] 1 SCR 623 [MMF]; Southwind v Canada, 2021 SCC 28 at paragraph 61, [2021] 2 SCR 450. Likewise, while the honour of the Crown permeates the relationship between the Crown and Indigenous peoples, it does not give rise to a freestanding cause of action: MMF at paragraph 73; Ontario (Attorney General) v Restoule, 2024 SCC 27 at paragraph 220 [Restoule]; Pekuakamiulnuatsh Takuhikan at paragraph 149. Replacing fiduciary duty with other private law doctrines, such as the tort of negligence, does not lead to a different result: the relationship is not a freestanding cause of action and it does not, alone, create sufficient proximity to ground a duty of care. Thus, the historical relationship does not result in the imposition on the federal government of a duty to protect Indigenous identity and culture whenever the latter are in jeopardy.

[146] The Plaintiffs rely on certain explicit commitments made in the course of the historical relationship, including treaties and the Manitoba Act, 1870. They argue that at their root, these commitments pertained to the protection of Indigenous ways of life or, in other words, culture and identity. This submission fails because it runs into the same problem of excessive generality. The duty that flows from it lacks any identifiable bounds. The Plaintiffs have not explained why this general commitment would translate into an all-encompassing duty of care and they did not point to any authority supporting this proposition. I emphasize that the Plaintiffs have not attempted to show a breach of treaty or of any other legal instrument.

(e) Federal Jurisdiction

[147] The Plaintiffs nevertheless submit that the federal government is in a relationship of proximity with class members because of Parliament’s jurisdiction in relation to Indigenous peoples. If I understand the argument correctly, federal jurisdiction would give rise to a duty to protect Indigenous culture and identity.

[148] In this regard, the Plaintiffs insist on the Supreme Court’s statement in Daniels, at paragraph 50, that “[n]on-status Indians and Métis are “Indians” under s. 91(24) and it is the federal government to whom they can turn.” This statement, however, does not have the scope that the Plaintiffs contend for. After Daniels, the federal government was no longer able to dismiss the claims of Métis and non-status Indians out of hand by asserting that Parliament lacked jurisdiction. The Court nevertheless stated that its judgment “does not create a duty to legislate” (at paragraph 15). Most importantly, Daniels did not retrospectively impose a duty of care on the federal government with respect to the acts or omissions of provincial governments.

[149] More generally, the allocation of jurisdiction to Parliament pursuant to section 91(24) is not sufficient to create proximity between the federal government and the class of persons over whom Parliament has jurisdiction. I am prepared to accept, for the sake of argument, that the protection of Indigenous peoples against local settlers was one of the purposes behind section 91(24), although the Supreme Court mentioned other purposes in Daniels, at paragraphs 25–26. This protective purpose, however, does not translate into an obligation to enact protective legislation or to monitor the manner in which the provinces were applying their own laws. Neither does it detract from the principles set out above, which prevent a duty of care from arising where the alleged harm is caused entirely by the actions or omissions of the other order of government. If the fiduciary relationship does not by itself give rise to a cause of action, it is difficult to understand why section 91(24) would.

[150] Moreover, the Plaintiffs do not seriously challenge the validity or applicability of provincial legislation pursuant to which class members were apprehended or placed for adoption. There is no doubt that provincial child welfare laws are valid and that they can apply to Indigenous persons: Natural Parents, at 773–774; NIL/TU,O Child and Family Services Society v BC Government and Service Employees’ Union, 2010 SCC 45 at paragraphs 38, 45, [2010] 2 SCR 696 [NIL/TU,O]; Bill C-92 Reference at paragraph 98.

[151] In spite of this, the Plaintiffs say that the federal government retains a role in the protection of Indigenous identity because the provinces lack constitutional authority in this regard. This, however, misstates the scope of provincial jurisdiction. It is true that provincial legislation cannot take away or impinge upon entitlements closely related to Indigenous identity (what has been described as “the core of Indianness”), such as Indian status: Natural Parents, at 777. Nevertheless, recent judicial decisions afford a wide margin of manoeuvre to the provinces for tailoring their legislation to the situation of Indigenous peoples, as long as the legislation can be linked to a provincial head of jurisdiction. In particular, most provinces have amended their child welfare legislation to make specific provisions regarding Indigenous children. The Supreme Court expressly recognized the validity of such provisions in NIL/TU,O, at paragraph 41.

[152] Lastly, the Plaintiffs rely on the “functional equivalency” between the manner in which status Indian children, on the one hand, and Métis and non-status Indian children, on the other hand, were treated by the child welfare system. This equivalency resulted from the fact that provincial legislation was applied to all children in the province, whether they had Indian status or not. I am also prepared to accept that provincial employees who applied the legislation did not always know the precise status of a child. Nonetheless, subject to the exceptions reviewed below, the federal government provided funds only with respect to children with Indian status. Thus, the factors that justified a finding of proximity in Brown are absent with respect to Métis and non‑status Indian children. The alleged functional equivalency does not detract from the fact that the federal government did not have any involvement in the apprehension, placement and adoption of class members.

(f) A Shared-Cost Program?

[153] At the hearing, the Plaintiffs put forward an additional basis for finding a duty of care. They asserted that provincial child welfare services for Métis and non-status Indians were in fact funded by the federal government through various forms of transfer payments, in particular those made pursuant to the Canada Assistance Plan, RSC 1985, c C-1. In other words, child welfare generally would be a “shared-cost program.” These transfer payments would constitute a sufficient federal intervention to justify a duty of care. As it was in fact funding child welfare services for Métis and non-status Indian children, the federal government would have been in a position to force the provinces to deliver services in a manner that would not jeopardize Indigenous identity and culture.

[154] I am unable to give effect to this last-minute submission. A shared-cost program does not give rise to a duty of care on the part of the federal government. Moreover, there is no evidence that child welfare was a shared-cost program.

(i) Shared-Cost Programs and Duty of Care

[155] Even though funding an activity may, in appropriate circumstances, be an indicium of proximity, the Plaintiffs’ submission must be assessed in the context of Canadian federalism. It is well known that over the last 75 years, the federal government has used its spending power to fund the provision of certain public services by the provinces, to ensure that citizens have access to similar services regardless of their province of residence. These programs are often called “shared-cost programs.”

[156] There is usually a minimal degree of federal intervention in the design of these programs: Finlay v Canada (Minister of Finance), [1993] 1 SCR 1080 at 1123–1124. For example, the main substantive condition set by subsection 6(2) of the Canada Assistance Plan is that eligibility for income assistance not be conditional on a period of residence in the province concerned. A decision regarding what substantive conditions to impose on the provinces would be a core policy decision that does not give rise to a duty of care.

[157] Moreover, there is no evidence that payments made pursuant to the Canada Assistance Plan had any role in prompting the provinces to apply their child welfare legislation to Métis and non-status Indian children. It is also unclear whether the federal government could withhold funding from a province for reasons not contemplated by the Canada Assistance Plan or by federal-provincial agreements made pursuant to it.

[158] I hasten to add that Brown did not deal with a shared-cost program. Shared-cost programs typically apply to all residents of a province, while the agreement in Brown pertained to status Indians only. Moreover, Brown dealt with a situation in which the province refused to provide the service in the absence of full federal funding.

[159] For these reasons, the fact that child welfare might have been a shared-cost program does not contribute to establishing proximity between the federal government and Métis and non‑status Indian children who were apprehended and placed in foster care or adopted pursuant to provincial legislation.

(ii) No Evidence That Child Welfare was a Shared-Cost Program

[160] In any event, the evidentiary record does not show that child welfare was in fact a shared‑cost program. It is well known that the main focus of the Canada Assistance Plan was income assistance, not child welfare or child protection. Income assistance is the provision of a basic income to those who are in need or unable to work and is colloquially known as “welfare.” Part I of the Canada Assistance Plan allows the federal government to enter into agreements with the provinces for the equal sharing of the costs of income assistance and related “welfare services.” Part II makes special provision for the full federal funding of income assistance and welfare services for status Indians living on reserves, again through the conclusion of federal‑provincial agreements. The Plaintiffs’ claim, however, has nothing to do with income assistance. Rather, it is based on a specific aspect of child welfare services, usually described as child protection, which involves the apprehension and placement of children whose well-being is jeopardized in their families, pursuant to legislation.

[161] It may be that some services associated with child welfare were eligible as “welfare services” pursuant to the Canada Assistance Plan. “Welfare services” aim at “the lessening, removal or prevention of the causes and effects of poverty, child neglect or dependence on public assistance” and include “adoption services” alongside items such as “rehabilitation services” and “homemaker, day-care and similar services.” Nevertheless, an overall reading of the statute suggests that child welfare and in particular child protection was not its main focus.

[162] The evidentiary record falls far short of showing on a balance of probabilities that child welfare for the general population was a shared-cost program during the relevant period. It was not assembled with this purpose in mind. Some documents contain passing references to the Canada Assistance Plan, but it is difficult to infer anything from them with any degree of confidence. In particular, a Treasury Board précis dated January 12, 1972 contains a proposal to enter into agreements with the provinces pursuant to part II of the Canada Assistance Plan, that is, with respect to services provided to status Indians. It would appear that the document lumps income assistance and child welfare together. It is therefore difficult to know whether certain statements regarding the sharing of costs with respect to the general population pertain to income assistance or child welfare. The record contains several federal-provincial agreements dealing with child welfare services for status Indians living on reserve, and none of them appear to have been made pursuant to part II of the Canada Assistance Plan, which suggests that child welfare services were outside the latter’s scope.

[163] The Plaintiffs rely heavily on the Defendant’s “admission,” in its statement of defence, that social programs eligible under part I of the Canada Assistance Plan “included child welfare services.” As explained above, there may have been some degree of overlap between child welfare services and the “welfare services” covered by part I. The Defendant’s “admission,” which merely restates the words of the statute, does not mean that child protection was a shared‑cost program.

[164] I asked the parties to provide post-hearing submissions on this issue, but they were not helpful. In fact, both parties overlooked a relatively recent Supreme Court precedent on the interpretation of the Canada Assistance Plan, Quebec (Attorney General) v Canada, 2011 SCC 11, [2011] 1 SCR 368. I do not need to rely on this case to buttress my findings above, but I do not see anything in it that would assist the Plaintiffs’ case.

[165] Thus, on the evidence before me, I am unable to find that the federal government funded the provision of child protection services for Métis and non-status Indian children through transfer payments. In any event, for the reasons stated above, this would not have established the required proximity to ground a duty of care.

(g) Summary Regarding Proximity with the Entire Class

[166] In summary, I acknowledge that there is a historical relationship between the federal government and Indigenous peoples. I also acknowledge that the Plaintiffs’ claim involves significant interests, namely, Indigenous culture and identity. However, subject to one exception discussed below, the federal government never intervened, directly or indirectly, in the provision of child welfare services to Métis and non-status Indian children. This, in my view, is determinative. There can be no proximity where the federal government simply did not intervene in the manner in which the provinces applied their own legislation to class members. Hence, the law of negligence does not require the federal government to compensate class members for the harms that have resulted from the application of provincial child welfare laws.

[167] Given the above, it is not necessary to address foreseeability in respect of the entire class.

C. Duty of Care With Respect to the Adopt Indian Métis Program

[168] Even though the federal government did not have a duty of care towards the entire class, it did have such a duty towards a subset of the class, namely, those children who were placed for adoption pursuant to the Government of Saskatchewan’s AIM program. This is because the federal government directly funded AIM, which, in the context of the historical relationship, creates the proximity necessary to establish a duty of care. The harm was foreseeable and there are no countervailing policy considerations negating such a duty.

[169] Of course, the Plaintiffs sought summary judgment in respect of the entire class, not only those children placed pursuant to AIM. Nevertheless, they invited the Court, if necessary, to give “nuanced and varied answers based on the situations of individual class members” (quoting from Vivendi Canada Inc v Dell’Aniello, 2014 SCC 1 at paragraph 46, [2014] 1 SCR 3). While the Defendant argued that there is no duty of care towards the whole class, it stated, in the alternative, that specific initiatives such as AIM could only establish proximity “with respect to particular Class Members affected by those local federal funding programs and only during the time periods in question.” Both parties filed extensive evidence concerning AIM, including more than 1000 pages of documents, and made specific submissions about the program. Contrary to what the Defendant suggested at the hearing, there is no unfairness in issuing judgment with respect to AIM only.

(1) Description

[170] In the fall of 1966, the Government of Saskatchewan applied to the federal Department of National Health and Welfare for a “Welfare Demonstration Grant” in respect of a “Special Adoption Unit to place Indian and Metis Children for Adoption.” The application noted the particular difficulty in finding parents willing to adopt Indigenous children. The proposed program would have three components. First, a general advertising component would seek to raise awareness about the possibility of adopting Indigenous children. Second, children ready for adoption would be advertised individually. Third, specialized staff would speed up the adoption process once a family expressed its interest.

[171] The Department of National Health and Welfare granted the funding request and AIM began its operations in early 1967. In a press release issued in May 1968, the Saskatchewan government stated that the number of Indigenous children who were adopted during AIM’s first year of operation was three times the number in the previous year. Two years after its inception, AIM produced a report in which it described its activities. During the first two years of operation, 110 children were placed for adoption, 49 of whom were Métis. The report compiled detailed data about the adoptive families, which indicated that nearly all adoptive fathers were non-Indigenous. Advertisements of individual children, sometimes using pseudonyms, were published in daily newspapers, and “one local television station advertised an individual child each week during the 6 p.m. newscast.” It added that “[t]he advertisement of individual children received an excellent response from the community.” While sample advertisements were not filed in evidence, Ms. Nora Cummings, a former president of the Saskatchewan Native Women’s Association, is seen holding two of them in the following picture found in the record (Plaintiffs’ Responding Motion Record at 108):

[172] The 1969 report concluded that AIM was a success and recommended its continuance. In particular, it suggested that “resources external to this province” should be considered for placing Indigenous children. Other pieces of evidence suggest that federal funding was AIM’s main, if not exclusive source of funding until at least 1972, with yearly amounts in the range of $35,000–$40,000. I understand that AIM was renamed REACH sometime after 1972, but there is little evidence in this regard. We do not know exactly how many Métis children were adopted pursuant to the AIM program overall.

(2) Proximity

[173] With respect to AIM, the issue of proximity presents itself in a fundamentally different manner compared to the entire class, because the federal government chose to fund a discrete program geared towards the adoption of Métis children (which also included children more properly identified as non-status Indian). It did not simply let the province apply its own legislation without intervening. Moreover, the consequences of the decision to fund AIM on the transmission of Indigenous identity and culture must be assessed in the context of the federal government’s historical relationship with Indigenous people. All of these factors militate towards a finding of proximity.

[174] When it funded the AIM program, the federal government knew that it would result in the permanent separation of Indigenous children from their families and communities. This was obvious from the grant application, and even from the name of the program itself, as adoption severs the relationship between a child and their biological parents. Moreover, there is every reason to believe that the federal government received the 1969 report regarding the program’s first two years of operation, as it provided most, if not all of the funding.

[175] It is true that AIM was operated by employees of the Saskatchewan government. Some of the program’s documents describe it as a joint federal-provincial initiative; others as a provincial program. These characterizations are immaterial. What matters is that funding the program enabled the harm it allegedly caused to class members.

[176] Of course, not every grant provided by the federal government creates proximity with persons affected by the program that receives the grant. In this case, however, the grant pertained to Indigenous peoples, with whom the federal Crown has a fiduciary relationship. Moreover, the servants of the federal Crown who approved the grant knew or must have known that the program would have serious impacts on the transmission of Indigenous identity and culture, which are interests of the utmost importance in this fiduciary relationship. In my view, this is sufficient to create proximity with the Métis and non-status Indian children who were placed and adopted pursuant to the AIM program. This also distinguishes the JB decision of the Ontario Court of Appeal, which did not involve a program targeting Indigenous children and that would have the obvious effect of hampering the transmission of Indigenous identity and culture.

[177] Hence, to paraphrase Lord Atkin in Donoghue at 580, the Métis and non-status Indian children placed pursuant to the AIM program were so closely and directly affected by what the federal grant enabled the Saskatchewan government to do that the federal government ought reasonably to have them in contemplation when deciding whether to fund the program or not.

[178] The absence of direct communication or personal relationship between the federal government and class members prior to the alleged negligent acts does not negate proximity. For example, in Hill v Hamilton-Wentworth Regional Police Services Board, 2007 SCC 41 at paragraph 29, [2007] 3 SCR 129, the Supreme Court found a duty of care between police officers and the person they were investigating, even though they had not been in contact before. In reality, in Donoghue, the phrase “close and direct” characterizes the effects of the defendant’s act on the plaintiff, not their prior relationships.

[179] Moreover, there is no indication that the grant was made pursuant to any statutory scheme. Hence, finding a duty of care does not affect the performance of statutory duties.

[180] In oral submissions, the Defendant characterized AIM as an outreach program that did not really form part of the machinery of child welfare. This is factually incorrect. While it is true that AIM included a component of general advertising that sought to change attitudes towards the adoption of Indigenous children, AIM employees were directly involved in the application of child welfare legislation to specific children, leading to their adoption. Indeed, Ms. Varley, one of the representative plaintiffs, was adopted through AIM. The evidence shows that employees of AIM were involved at several stages of the adoption process. While the evidence does not disclose whether she was advertised in the newspapers, the case notes include a mention that pictures of her were taken and forwarded to AIM. It is obvious that AIM’s involvement contributed to the harm alleged in the statement of claim.

[181] I would simply add that proximity does not depend on the identity categories used to describe a child at the time of placement and adoption nor on the manner in which they identify today. For example, Ms. Varley was described as Métis in the case notes, while she now identifies as non-status Indian. What matters is that AIM was a program for Indigenous children irrespective of whether they held Indian status or not.

(3) Foreseeability

[182] To establish a duty of care, one must also prove that “the risk of the type of damage that occurred was reasonably foreseeable to the class of plaintiff that was damaged”: Rankin (Rankin’s Garage & Sales) v JJ, 2018 SCC 19 at paragraph 24, [2018] 1 SCR 587 [Emphasis in original]; see also Imperial Tobacco at paragraph 57. Here, it was reasonably foreseeable that funding AIM would sever the connection between children adopted through that program and their Indigenous family, community and culture. This is largely a matter of common sense. Culture is transmitted by processes of socialization, first through the family, then through community institutions such as schools: Côté at paragraph 56; Quebec Reference at paragraph 58; Bill C-92 Reference at paragraph 113. Indigenous children who are removed from their families and communities at a young age, to be placed and eventually adopted in non-Indigenous families, will most likely be disconnected from their Indigenous culture, giving rise to the type of harms alleged in the statement of claim.

[183] There is every indication that this was well understood during the period covered by the claim. In the 1975 Natural Parents case, the Supreme Court quoted the following excerpt of the trial judge’s reasons (at 768):

Those who gave evidence, as well as the Court’s own advisers, were all of the opinion that there was potential danger to a native child being brought up in a white family, particularly when he reached the later stages of adolescence. I can readily appreciate this view: it is based on perfectly sound ideas of the effects of heredity and is not a matter merely emotional or racial. Instances abound where such persons have in the past experienced difficulty in establishing racial identity in their maturity.

[Emphasis added]

[184] Likewise, in 1982, a summary of recommendations regarding the renewal of the agreement with Manitoba shows that the impacts of child welfare legislation on the transmission of Indigenous culture were well understood:

The family is the first resource for the nurture and protection of children but families do need support for their parenting role and children, for a variety of reasons, may need substitute care . . .

As a result of culture, geography and experiential past, Indian people have special needs;

Preservation of Indian cultural identity is of a paramount importance, in terms of both language and customs, within the framework of tribes, bands, communities, extended families and individuals . . .

(4) Residual Policy Considerations

[185] Proximity and foreseeability establish a prima facie duty of care. The Defendant then has the burden of showing that residual policy considerations negate this duty of care: Marchi at paragraph 35. In my view, the Defendant has not discharged this burden.

(a) Not a Core Policy Decision

[186] First, the Defendant has not shown that the decision to fund AIM was a core policy decision according to the criteria laid out in Marchi. The evidence contains a general description of the grant program pursuant to which AIM was funded. Although the evidence is silent in this regard, one must presume that there were objective criteria by which proposals were judged. The yearly amount of the grant, relative to a department’s budget, is modest. There is no indication that budgetary considerations were relevant in deciding whether to accede to Saskatchewan’s request. While the letter awarding the grant is signed by the Deputy Minister, one must assume that the proposal was reviewed by civil servants in the Department of National Health and Welfare. In contrast to the general policies reviewed above, there is no evidence that the decision to fund AIM was discussed among several departments or that it resulted from a balancing of competing policy considerations. These are all indications that the decision in question was operational in nature.

[187] The fact that the claim of the entire class is dismissed because of the immunity for core policy decisions has no bearing on the liability in respect of the AIM program. The core policy decision in question was the decision not to fund services for Métis and non-status Indians where the provinces already offer those services to the general population. As a matter of fact, this policy did not prevent the federal government from funding AIM.

(b) Not a Duty to Legislate

[188] The Defendant argued that finding a duty of care in this case would amount to a duty to legislate, because an amendment to the Indian Act or new legislation would be required to identify the members of the class. The facts, however, do not bear this out.

[189] In reality, AIM was able to operate without any statutory definition of Métis or non-status Indians. Admittedly, social workers applying provincial legislation may have had difficulty ascertaining whether a particular child had Indian status or may have used the wrong identity category. It may be that at the time, Saskatchewan government officials used the term “Métis” as encompassing non-status Indians. Nevertheless, there is no indication that they had any difficulty in identifying Indigenous children. The 1969 report does not mention any concern in this respect.

[190] More generally, it is possible to design legislation or government programs aimed at Indigenous peoples without a precise definition of status. For example, the principles of sentencing established in R v Gladue, [1999] 1 SCR 688, apply to all Indigenous persons irrespective of status. The child welfare legislation of certain provinces, such as Ontario’s Child, Youth and Family Services Act, 2017, SO 2017, c 14, Sch 1, applies to Indigenous children irrespective of status. Bill C-92 does the same.

(c) Not an Indeterminate Class

[191] The Defendant also argued that finding a duty of care in this case would lead to indeterminate liability towards an indeterminate class. It said that the federal government did not possess lists of Métis or non-status Indians and that it would have been impossible, at the time, to ascertain who the members of the class were.

[192] In my view, these concerns are overblown. A duty of care may arise even though the defendant does not know the identity of the members of the class to whom it owes the duty. This is almost invariably the case in product liability cases, such as Donoghue.

[193] I acknowledge that identifying members of the entire class certified by Justice Phelan could give rise to difficulties. These difficulties, however, do not arise with respect to children adopted through AIM. The identity of these children is ascertainable from existing records, as in Ms. Varley’s case. One can assume that those children were all Indigenous, given the focus of the program.

[194] To summarize, there are no residual policy considerations that would negate the duty of care towards children who were placed or adopted through the AIM program.

[195] Hence, as proximity and foreseeability are established, the federal government had a duty of care towards class members who were placed or adopted through the AIM program.

(5) Servants of the Crown

[196] As this claim is governed by section 3 of the Crown Liability and Proceedings Act, the duty of care must pertain to acts or omissions of servants of the federal Crown. I have no difficulty finding that the management of the federal grant pertaining to AIM was in the hands of servants of the Crown, even though they are not all identified by name. The letter awarding the grant was signed by the Deputy Minister of National Health and Welfare, and one must assume that the grant was managed by servants of the Crown working in that Department.

D. Fiduciary Duty

[197] The Plaintiffs also assert that the federal government has a fiduciary duty towards class members. In this regard, one must keep in mind that although the relationship between the Crown and Indigenous people is fiduciary in nature, not all aspects of this relationship give rise to legally enforceable fiduciary duties: Wewaykum at paragraph 83; Restoule at paragraphs 241–242.

[198] In MMF, at paragraphs 49–50, and in Restoule, at paragraph 222, the Supreme Court stated that a fiduciary duty may arise in two ways. First, in the Indigenous context, it may result from the Crown assuming discretionary control over specific Indigenous interests. This is the “sui generis” fiduciary duty. Second, an “ad hoc” fiduciary duty may arise from an undertaking, if the criteria set forth in Alberta v Elder Advocates of Alberta Society, 2011 SCC 24 at paragraph 36, [2011] 2 SCR 261, are met. These criteria are:

(1) an undertaking by the alleged fiduciary to act in the best interests of the alleged beneficiary or beneficiaries; (2) a defined person or class of persons vulnerable to a fiduciary’s control (the beneficiary or beneficiaries); and (3) a legal or substantial practical interest of the beneficiary or beneficiaries that stands to be adversely affected by the alleged fiduciary’s exercise of discretion or control

(1) Ad Hoc Fiduciary Duty

[199] No ad hoc duty arose in this case. Firstly, the Plaintiffs did not point to any undertaking of the federal government to act in the best interests of Métis and non-status Indian children with respect to child welfare services. Rather, the evidence shows that with rare exceptions, the federal government did not undertake to provide services to Métis and non-status Indian children, let alone to put their interests ahead of those of others.

[200] Secondly, during the period covered by the action, the federal government did not exercise any discretionary power over Indigenous identity and culture, even assuming that the latter are a “legal or substantial practical interest of the beneficiary.” It did not take any decision regarding a child’s apprehension, placement or adoption; this was done by provincial officials. This case is unlike Paddy-Cannon v Attorney General (Canada), 2023 ONSC 6748, in which federal officials actively intervened in a child welfare case.

(2) Sui Generis Fiduciary Duty

[201] The Plaintiffs argue that the federal government owes a sui generis fiduciary duty to class members because it assumed discretionary control over “the preservation of the cultural ties and the prevention of cultural assimilation.”

[202] Most cases in which a sui generis fiduciary duty was found to exist related to the management of reserve land. See, for example, Guerin v The Queen, [1984] 2 SCR 335; Blueberry River Indian Band v Canada (Department of Indian Affairs and Northern Development), [1995] 4 SCR 344; Wewaykum; Williams Lake Indian Band v Canada (Aboriginal Affairs and Northern Development), 2018 SCC 4, [2018] 1 SCR 83. The Crown is heavily involved in the process of reserve creation, management and surrender. The Indian Act vests it with considerable discretionary power and First Nations are vulnerable to decisions made by the Crown in this regard.

[203] Putting the Plaintiffs’ case at its highest, I am prepared to assume that the Indigenous interests that may be the subject of a sui generis fiduciary duty extend beyond reserve lands and may include the protection of Indigenous identity and culture and that these interests are protected by section 35 of the Constitution Act, 1982. Nevertheless, for a fiduciary duty to arise, the Crown must have assumed discretionary control over the interest in question. As the Supreme Court wrote in Wewaykum, at paragraph 83, the inquiry must “focus on the particular obligation or interest that is the subject matter of the particular dispute and whether or not the Crown had assumed discretionary control in relation thereto sufficient to ground a fiduciary obligation.”

[204] The federal government assumed no such discretionary control during the period covered by the action in relation to child welfare services. These services were administered by provincial officials, under statutory schemes that typically required a court order to approve major decisions. The federal government did not have any power to approve a child’s apprehension, placement or adoption that would be similar to its power to approve or reject a proposal for the surrender of reserve land. While it is often said that the federal government had a “high degree of control” over the lives of Indigenous peoples, generally because of paternalistic provisions of the Indian Act, it did not have any control in respect to child welfare in respect of members of the class.

[205] Contrary to the intervener Manitoba Métis Federation’s submissions, the federal government’s funding of the provinces in respect of child welfare for status Indians did not amount to an exercise of jurisdiction or an assumption of discretionary power over Métis and non-status Indian children. The federal government could not oversee the application of provincial legislation to the latter children, nor could it impose conditions, cancel funding or require consultation with Indigenous communities. As I explained above, the statutory framework for transfer payments, for example payments made pursuant to the Canada Assistance Plan, does not grant the federal government any discretionary power beyond highly general policy decisions that would attract immunity.

[206] Hence, the federal government did not owe a legally enforceable sui generis fiduciary duty towards members of the class in relation to the preservation of their culture and identity.

V. Disposition

[207] The Defendant’s motion for summary judgment will be dismissed because I cannot determine, on a class-wide basis, that the class members’ monetary claims are time-barred pursuant to section 32 of the Crown Liability and Proceedings Act.

[208] The Plaintiffs’ motion for summary judgment will be granted only to the extent that I have found that servants of the federal Crown owed a duty of care towards class members who were placed in foster care or adopted pursuant to the AIM program. As I have found that the federal government did not owe a duty of care to the entire class nor any fiduciary duty, the Plaintiffs’ motion will be dismissed in all other respects.

ORDER in T-2166-18

THIS COURT ORDERS that:

The Defendant’s motion for summary judgment is dismissed.
The Plaintiffs’ motion for summary judgment is granted in part.
Servants of His Majesty the King in right of Canada had a duty of care towards class members who were placed in foster care or adopted through the Adopt Indian Métis or AIM program of the Government of Saskatchewan.
The Plaintiffs’ motion for summary judgment is dismissed in all other respects.

"Sébastien Grammond"

Judge

FEDERAL COURT

SOLICITORS OF RECORD

	Docket:	T-2166-18
STYLE OF CAUSE:		SHANNON VARLEY, SANDRA LUKOWICH v THE ATTORNEY GENERAL OF CANADA
PLACE OF HEARING:		BY VIDEOCONFERENCE
DATE OF HEARING:		December 9–11, 2024
order and REASONS:		GRAMMOND J.
DATED:		april 29, 2025

APPEARANCES:

Celeste Poltak

Andrew Lokan

Kirk M. Baert

Jamie Shilton

Zara Narain

Linda Rothstein

Lindsay Scott

For The Plaintiffs

Craig A.B. Ferris, K.C.

Marko Vesely, K.C.

Laura E. Duke

For The Defendant

John Carlo Mastrangelo

Rahool P. Agarwal

Jenna D’Aurizio

FOR THE INTERVENER

MANITOBA MÉTIS FEDERATION

Jason T. Madden

Alexandria J. Winterburn

Steve Tenai

Matthew Patterson

FOR THE INTERVENERS

OTIPEMISIWAK MÉTIS GOVERNMENT

MÉTIS NATION OF ALBERTA ASSOCIATION

SOLICITORS OF RECORD:

Koskie Minsky LLP

Toronto, Ontario

Paliare Roland Rosenberg Rothstein LLP

Toronto, Ontario

For The Plaintiffs

Lawson Lundell LLP

Vancouver, British Columbia

For The Defendant

Lax O’Sullivan Lisus Gottlieb LLP

Toronto, Ontario

FOR THE INTERVENER

MANITOBA MÉTIS FEDERATION

Aird & Berlis LLP

Toronto, Ontario

FOR THE INTERVENERS

OTIPEMISIWAK MÉTIS GOVERNMENT

MÉTIS NATION OF ALBERTA ASSOCIATION

You are being directed to the most recent version of the statute which may not be the version considered at the time of the judgment.

Do not display this message anymore