No. 25-1836

Appeal from the United States District Court for the Eastern District of Michigan at Detroit. No. 2:24-cv-13195 — David M. Lawson, District Judge.

Argued: July 30, 2026

Decided and Filed: August 28, 2026

Before: STRANCH, BUSH, and MURPHY, Circuit Judges.

ARGUED: Anita M. Washington, BEY & ASSOCIATES, LLC, Cincinnati, Ohio, for Appellant. Cynthia M. Filipovich, CLARK HILL PLC, Detroit, Michigan, for Appellee. ON BRIEF: Carla D. Aikens, CARLA D. AIKENS, P.L.C., Detroit, Michigan, for Appellant. Cynthia M. Filipovich, Maria Fracassa Dwyer, CLARK HILL PLC, Detroit, Michigan, for Appellee.

JOHN K. BUSH, Circuit Judge. As a general rule, nonparties to a litigation cannot be bound to a judgment. But, as with most general rules, this rule has exceptions — six, in fact. In this appeal, we consider whether one of those exceptions prevents a plaintiff from suing a Page 2corporation’s wholly owned subsidiary after unsuccessfully suing the parent corporation on the same claim. The subsidiary has not shown that any exception applies here, so we REVERSE .

Diamond Williams brought this suit alleging that she experienced race and gender discrimination, harassment, and retaliation while employed at a Mastronardi Produce-USA (USA) facility in Livonia, Michigan. This was the second lawsuit Williams filed on these facts. In the first, she sued Mastronardi Produce, Ltd. (Canada), the parent company of USA, claiming that Canada was her employer and had discriminated against her. See Williams v. Mastronardi Produce, Ltd. , No. 23-13302, 2024 WL 4932716 (E.D. Mich. Dec. 2, 2024). Canada defended by moving to compel arbitration and dismiss the action, asserting as a factual matter that Williams “sued the wrong entity.” Id. at *1. The district court denied both the motion to compel and the motion to dismiss and allowed the suit to proceed. Id. at *2.

After the ruling, Canada's counsel emailed Williams's counsel, explaining that the lawsuit named the wrong entity and asking whether she would agree to substitute USA as the defendant. Id. If Williams would not agree to substitute, Canada would seek judgment on the pleadings. Id. In the interim, Mastronardi answered the complaint and attached a declaration from its Senior Director of Human Resources explaining that Canada and USA are separate entities with separate policies governing their separate workforces, and none of the persons mentioned in the complaint have ever been Canada employees. Id.

Williams's counsel refused to substitute USA as the defendant, so Canada moved for judgment on the pleadings. Id. Williams argued in opposition that she was employed by Canada, and, alternatively, conditionally requested leave to amend to substitute the proper defendant. Id. at *3. The district court rejected the 'conditional' response as procedurally improper because Williams needed to either seek consent from Canada or seek unconditional leave of court to amend her complaint. See id. Williams filed another conditional motion in response, which the district court denied for failing to follow the local rules. Id. At the same time, the district court converted the motion for judgment on the pleadings into a motion for summary judgment and ordered supplemental briefing. Id. Williams did not renew her motion Page 3to amend but filed a supplemental brief attaching two deposition excerpts suggesting that some employees of either USA or Canada were uncertain who their employer was. Id. In Canada’s supplemental briefing, it provided W2 statements showing USA as Williams’s employer. Id. The district court sided with Canada because the evidence established that Canada was not Williams’s employer, the complaint did not plead a joint -employer theory of liability, and any attempt to pierce the corporate veil based on Canada's ownership of USA would fail on the record presented. Id. at *5 – 7.

Shortly before the opinion was issued, Williams sued USA asserting nearly identical factual allegations to those underlying her earlier case, with the addition of a claim for “Hostile Workplace Environment” under Michigan law. USA moved to dismiss the second case under Federal Rule of Civil Procedure 12(b)(6), arguing that claim preclusion barred the suit because USA and Canada are privies. As part of this argument, USA asserted privity based on its control over Canada’s defense in the prior litigation, including by sharing defense counsel, offering to substitute as the proper defendant, providing evidence to establish that it was Williams's employer, and submitting a declaration identifying the individuals named in the complaint as USA employees. The district court rejected this basis for privity and found that the record did not support the claim that USA controlled the prior litigation.

Instead, the district court found privity between USA and Canada by relying on the closeand-significant-relationship test used by some of our sister circuits. It based this finding on USA and Canada’s parent -subsidiary relationship, as well as equitable considerations that Williams had not established a sufficient reason to receive 'a second bite at the apple.' R. 11, Order, PageID 329 – 331. Because the district court concluded that USA and Canada were privies and that the remaining elements of claim preclusion were satisfied, the district court granted USA’s motion to dismiss.

Page 4Williams now brings this timely appeal claiming that the district court erred by adopting the close-and-significant-relationship test to conclude that USA and Mastronardi Canada were privies.¹

'The preclusive effect of a federal -court judgment is determined by federal common law.' Taylor v. Sturgell , 553 U.S. 880, 891 (2008). Although the existence of privity and control are questions of fact, Vulcan, Inc. v. Fordees Corp. , 658 F.2d 1106, 1109 (6th Cir. 1981), we 'review de novo a district court’s application of the doctrine of res judicata,” United States ex rel. Sheldon v. Kettering Health Network , 816 F.3d 399, 407 (6th Cir. 2016). We also review “ de novo a district court’s dismissal of a suit pursuant to Rule 12(b)(6).” Id. Claim preclusion is an affirmative defense. See Chapman v. JPMorgan Chase Bank, N.A. , 651 F. App’x 508, 513 (6th Cir. 2016). And a plaintiff's complaint generally 'need not plead factual allegations to plausibly avoid an affirmative defense.” VCST Int’l B.V. v. BorgWarner Noblesville, LLC , 142 F.4th 393, 399 – 400 (6th Cir. 2025). But a district court may grant a motion to dismiss based on claim preclusion if the complaint’s factual allegations (when combined with the public records of the prior suit that we may consider at this stage) 'conclusively establish' that this preclusion applies. Chapman , 651 F. App'x at 513 (citation omitted); see Stevens v. St. Tammany Par. Gov’t , 17 F.4th 563, 571 (5th Cir. 2021); Parungao v. Cmty. Health Sys., Inc. , 858 F.3d 452, 457 (7th Cir. 2017).

Claim preclusion bars 'successive litigation of the very same claim' when a final judgment has been entered. New Hampshire v. Maine , 532 U.S. 742, 748 (2001). This prevents parties from relitigating “matters that they have had a full and fair opportunity to litigate . . . .”Page 5 Taylor , 553 U.S. at 892 (quoting Montana v. United States , 440 U.S. 147, 153 (1979)). Nonparties to the first litigation, on the other hand, are generally not bound to the outcome because they did not have a “‘full and fair opportunity to litigate’ the claims and issues settled in that suit.” Id. (quoting Montana , 440 U.S. at 153). But this rule is subject to six exceptions that justify precluding a nonparty from relitigating a claim or issue. Id. at 893.²

Our preTaylor case law grouped multiple of these exceptions into a single privity exception. A party was a privy if it was (1) a successor in interest, (2) 'a nonparty who controlled the original suit,” or (3) “a nonparty who is adequately represented by a party . . . .” Becherer v. Merrill Lynch, Pierce, Fenner & Smith, Inc. , 193 F.3d 415, 422 (6th Cir. 1999) (en banc). In Taylor , the Supreme Court moved the analysis toward more precise categories, albeit without providing a “definitive taxonomy” of the various exceptions. 553 U.S. at 893 n.6. It recognized that many lower courts were using the term 'privity' to refer to several of the exceptions without distinguishing among them. Id. at 894 n.8. So Taylor avoided the term privity altogether and focused instead on whether one of the six common-law exceptions would allow for non-party claim preclusion. Id. at 893 – 95 & n.8. PostTaylor , we now ask: does the case fit into one of Taylor ’s six exceptions that justify non -party claim preclusion?

Not every circuit has agreed with this reading of Taylor or our framing of the question. See Airframe Sys., Inc. v. Raytheon Co. , 601 F.3d 9, 17 & n.8 (1st Cir. 2010) (limiting Taylor to defensive uses of non-mutual claim preclusion against non-party plaintiffs and allowing preclusion based on a close and significant relationship with the prior defendant); Elbert v. Carter , 903 F.3d 779, 783 (8th Cir. 2018) (applying preTaylor circuit precedents to conclude that privity barred a suit because the new and old defendants had a close relationship). These readings of Taylor are unpersuasive.

Page 6The First Circuit's approach distinguishes Taylor based on the facts, but not the reasoning. But 'just as binding as [the] holding is the reasoning underlying it.' Bucklew v. Precythe , 587 U.S. 119, 136 (2019). When Taylor laid out the six categories of exceptions, it did so without reference to which side of the “v.” the party is on, and its reasoning applies with equal force to defendants as to plaintiffs. See 553 U.S. at 893 ('Though hardly in doubt, the rule against nonparty preclusion is subject to exceptions. For present purposes, the recognized exceptions can be grouped into six categories.”). The Eighth Circuit likewise failed to root its rule in Taylor ’s reasoning; it notes only that “ Taylor involved different considerations than this case.” Elbert , 903 F.3d at 784. As Judge Kelly noted in her separate writing, the majority in Elbert invoked privity in exactly the manner that Taylor rejected. Id. at 785 – 86 (Kelly, J., concurring in part and dissenting in part).

The district court adopted the approach of the First and Eighth Circuits. That was error. Regardless of the supposed practical benefits of a looser standard, the Supreme Court has given clear guidance for when an exception to non-party preclusion is appropriate. So rather than look to whether the parties have a close and significant relationship, we instead ask whether the parties fit into one of the traditional exceptions.

Three exceptions are at issue: (1) the party asserting preclusion has a pre-existing substantive legal relationship with a party to the first litigation, (2) the nonparty assumed control of the prior litigation, and (3) the nonparty’s interests were “adequately represented by someone with the same interests who was a party to the suit.” Taylor , 553 U.S. at 894 – 95 (cleaned up). We take each in turn.

The first exception for pre-existing substantive legal relationships cannot justify claim preclusion here. This is the exception that looks most like common-law privity. And it was this common-law tradition that the Supreme Court cited as its basis for recognizing this exception. Id. at 894. So we ask two questions: First , what was the common law tradition relating to special legal relationships? Second , do the parties here fit into that tradition?

Page 7Start with the first question. At common law in England, res judicata generally bound only the parties to the action and those with specific legal relationships to those parties. So the rule was that judgments were binding on only the immediate parties and their privies. See 2 Edward Coke, The First Part of the Institutes of the Laws of England 352a -352b (1st Am. ed. 1812). This concept of privity that justified preclusion was narrower than what our circuit laid out in Becherer . Privity at common law denoted a 'mutual or successive relationship to the same rights of property.” Simon Greenleaf, A Treatise on the Law of Evidence 216 (12th ed. 1866); Bigelow v. Old Dominion Copper Mining & Smelting Co. , 225 U.S. 111, 128 – 29 (1912). These mutual or successive relationships could arise through inheritance, succession, or the transfer of property interests. See 2 William Blackstone, Commentaries *355 – 56.

The United States imported this common law concept into American law. Early American courts recognized three types of privies that could be bound: privies in law, privies in blood, and privies in estate. A.C. Freeman, A Treatise on the Law of Judgments: Including All Final Determinations of the Rights of Parties in Actions or Proceedings at Law or in Equity 177 (5th ed. 1925). But when it comes to the binding effects of judgments, all privies are affected only as privies in estate. “They are bound because they have succeeded to some estate or interest which was bound in the hands of its former owner; and the extent of the estoppel, so far as the privy is concerned, is limited to controversies affecting this estate or interest.” Id. Most often, privity could be found in specific legal relationships like those between administrators and executors, heirs and devisees, principals and agents, assignees and assignors, bailors and bailees, garnishors and garnishees, heirs and ancestors, lessors and lessees, and successors and predecessors in office. See id. at 178 – 86; Bigelow , 225 U.S. at 127 – 29. In other words, the common-law practice was to find privity based on a successive relationship to property.

Having now addressed common-law practice, we can move to the second question: does the parent-subsidiary relationship between USA and Canada fit into this tradition? We answer no for two reasons.

First , the parent-subsidiary relationship between USA and Canada does not create an identical interest in property such that the property right was already adjudicated in the first litigation. The common law property-based relationships that justified preclusion did so because Page 8the property interest was conclusively determined in the first litigation. If A sued B regarding the ownership of Blackacre and A was found to be the owner, A would not need to relitigate the ownership interest against B ’s heir because B ’s heir has exactly the same legal interest as B . So the first litigation already determined the exact interest at issue in a future litigation against the heir. When two parties have the same interest, the interests are “alike concluded” by the first litigation. Bigelow , 225 U.S. at 129. This explains why nonmutual preclusion was propertybased at common law — the interests were always identical or representative such that the party in the first action had the same interest or stood in the shoes of the nonparty.

Parent companies and their wholly owned subsidiaries do not fit this bill. Each owns separate property, claims, liabilities, and contracts — that is, after all, one of the major reasons to use the corporate form. See Dole Food Co. v. Patrickson , 538 U.S. 468, 474 – 75 (2008); Galette v. N.J. Transit Corp. , 607 U.S. 509, 524 (2026). In individual cases, these property interests could overlap where the companies are joint owners of some property, but co-owners of property were not bound by judgments against the other. Co-owners have distinct interests in property, not successive ones. See Freeman, supra , at 187. This distinction is similar to the distinction between joint tenants and tenants in common. Tenants in common each own separate shares of the property and these shares are passed separately to the tenants’ respective heirs. United States v. Craft , 535 U.S. 274, 279 – 80 (2002). A tenancy in common does not include a right of succession from the cotenant’s interest. Id. at 280. But two joint tenants have exactly the same remainder interest because they each have a right of survivorship. Id. So any adjudication affecting the remainder interest of one joint tenant would conclusively decide the remainder interest for both of them without going through a second litigation. That is not the case with tenants in common. So any co-ownership without a right of survivorship would include interests that are distinct, divisible, and pass separately to the coowners’ respective heirs.

Second , even if the parties did have a substantive legal relationship giving them identical interests in some property, that property-based connection was not at issue in the prior litigation nor here. Williams alleges that USA and Canada were joint tortfeasors who discriminated against her. And at common law, joint tortfeasors were not in privity for res judicata purposes. Bigelow , 225 U.S. at 129 (“[T]he sounder reason, as well as the weight of authority, is that the Page 9failure to recover against one of two joint tort feasors is not a bar to a suit against the other upon the same facts.”). It may help to flesh this out with an example. Say that a father and son were out hunting together and one of them accidentally shot a passerby. The passerby survived and sued the father believing that he was the one who fired the shot. As discovery proceeded, it became clear that the father’s gun did not work properly and he could not have fired the shot. So the passerby sues the son in a separate suit. Should this second suit be barred by claim preclusion simply because the father and son happen to be privies when it comes to property ownership? No. Precluding the passerby’s claim based on the relationship between the father and son would find no support in the common law or in common sense. The property that makes the parties privies must also be at issue in the case. See Restatement (Second) of Judgments § 43 (Am. L. Inst. 1982); 18A Wright & Miller’s Federal Practice & Procedure § 4448 (3d ed. Apr. 2026 Update).

Canada owning all the shares in USA does not relate to the alleged employment discrimination claim, so any hypothetical successive property relationship does not bar the suit. Like the father and son, USA and Canada each own their own tort claims and liabilities. Any successive property relationship between the two does not put them on the hook for each other’s tort liabilities.

Thus, claim preclusion cannot be justified based on USA and Canada's pre -existing substantive legal relationship.

B.

The control exception likewise fails. To determine whether a nonparty exercised sufficient control over the prior litigation, this circuit “requires that [the nonparty] have effective choice as to the legal theories and proofs to be advanced on behalf of the party to the action. [It] must also have control over the opportunity to obtain review.' Becherer , 193 F.3d at 423 (quoting Benson & Ford, Inc. v. Wanda Petroleum Co. , 833 F.2d 1172, 1174 (5th Cir. 1987)). Control is a high bar and requires a relationship between the nonparty and the controlled party akin to president-shareholder, parent-subsidiary, or liability insurer and its insured. Id. But the nature of the relationship is not sufficient to establish control. Id. Neither are isolated steps like Page 10retaining a lawyer, signing affidavits, contributing money to the first case, and communicating regarding the progress of the case. Id. Control is a question of fact, Leaf v. Refn , 742 F. App’x 917, 925 (6th Cir. 2018), so USA must show that Williams’s complaint (along with the properly reviewable “records of [her] earlier lawsuit”) leave no doubt that USA controlled the litigation, Chapman , 651 F. App’x at 513.

USA points to several facts to show its control over the prior litigation: the parentsubsidiary relationship, shared counsel between USA and Canada, USA's agreement to substitute as a defendant, and a shared corporate officer of USA and Canada filing a declaration and advising the court that the alleged wrongdoers were employees of USA. These facts do not clear the high bar of control set out in Becherer . Each of the facts USA points to could just as easily show that Canada exercised control over USA. Canada and USA had the same counsel and a shared officer testified. The record just as easily suggests that Canada chose the counsel and ordered the employee to testify. Canada also could have required USA, as its subsidiary, to offer to substitute as the defendant so that Canada would be off the hook. Nothing in the record establishes that these were USA's choices rather than Canada's. Instead, the facts suggest coordination — not control. See Becherer , 193 F.3d at 423; Benson & Ford , 833 F.2d at 1174 – 75. The parent-subsidiary relationship also does not tend to show control here because USA is arguing that the subsidiary exercised control over the parent, not the other way around. Because of this, the parent-subsidiary relationship cuts against control here, not for it.

The district court did not err in finding that the complaint failed to plead allegations suggesting that USA controlled the prior litigation.

The adequate-representation exception also cannot justify preclusion. Amos v. PPG Industries, Inc. , 699 F.3d 448 (6th Cir. 2012), decides this issue and the district court erred by not following that decision. The district court held that Canada had adequately represented USA’s interests in the prior litigation such that preclusion could apply. It based this holding on three reasons: (1) a close and significant relationship between USA and Canada, (2) USA should have been joined in the first action, and (3) Williams did not offer good reasons that justified Page 11giving her a “second bite at the apple.” R. 11, Dist. Ct. Op., PageID 323– 32. Once again, the district court applied the wrong test. Looking at the test laid out by our case law, none of these reasons matter.

The adequate-representation exception always requires both (1) an alignment of interests between the party and nonparty, and (2) “special procedures to protect the nonparties’ interests or an understanding that” a party to the first suit had acted as a “representative” to the nonparties. Amos , 699 F.3d at 452 (quoting Taylor , 553 U.S. at 897); see 202 N. Monroe, LLC v. Sower , 850 F.3d 265, 274 –75 (6th Cir. 2017). It also “sometimes” requires a third element: “notice of the original suit” to the nonparties. Amos , 699 F.3d at 452 (quoting Taylor , 553 U.S. at 900). In Amos , the court made quick work of the exception. It was not met because the district court 'neither certified a class (nobody asked it to) nor employed any other 'special procedures' to protect the [nonparties’] interests in that action.” Id. And no proof of assent to litigating in a representative capacity was presented by the party seeking preclusion. Id. at 453.

The same is true here. This case is not a class action and USA, the party seeking preclusion, did not point to any special procedures used by the district court in the first case to protect its interests. It also did not present evidence that Canada was litigating the first case in a representative capacity. Because USA satisfied neither part of the third element, the adequaterepresentation exception cannot preclude Williams from bringing this suit.

USA asserted at oral argument that even if claim preclusion does not bar the suit, Williams’s suit is barred by issue preclusion. This point was not raised before the district court, and any argument on this issue would be best left to the district court in the first instance. See PCC Airfoils, LLC v. Daughtery , 176 F.4th 509, 513 (6th Cir. 2026).

That said, we are not sure how issue preclusion could bar the entire suit. USA seems to misunderstand what effect the prior determination would have on the present litigation. The issue established in the prior litigation was that Canada is not Williams’s employer. USA argues now that issue preclusion bars Williams from asserting anything besides her prior position that Canada is her employer. See D. 35, Rule 28(j) Letter, at 2 (“Plaintiff nonetheless maintained that Page 12the named defendant Mastronardi Canada was her employer. Plaintiff is now precluded from arguing otherwise in the Subsequent Action.'). Under USA's view, it appears that a party is obligated to continue pressing a position even after it has been definitively rejected. But the exact opposite is true. If issue preclusion bars Williams from taking any position, it bars her from asserting that Canada is her employer because that is the issue that was conclusively litigated in the prior case. So if she is precluded at all, it is from asserting a position inconsistent with the prior ruling. See Ga.-Pac. Consumer Prods. LP v. Four-U-Packaging, Inc. , 701 F.3d 1093, 1103 (6th Cir. 2012) (“One of our primary goals in applying issue preclusion is to ‘foster[] reliance on judicial action by minimizing the possibility of inconsistent decisions.'' (quoting Montana , 440 U.S. at 153 – 54)).

But the record is not developed enough on this point, and we trust the district court to ably sort through these issues in the first instance if the parties raise them.

For the reasons outlined above, we REVERSE the decision of the district court.