Police-Coerced Witness Accounts as Actionable Fabrication and Brady Material—Qualified Immunity Denied Despite Prosecutorial Use of the Statements

Case: George Clark v. Anthony Abdallah (with co-plaintiff Kevin Harrington)
Court: United States Court of Appeals for the Sixth Circuit
Date: March 13, 2025
Posture: Interlocutory appeal from denial of qualified immunity at summary judgment in a 42 U.S.C. § 1983 action.

I. Introduction

This decision addresses how far qualified immunity shields detectives accused of (a) coercing key witnesses into inculpatory accounts, (b) initiating and maintaining prosecutions allegedly lacking probable cause once the coerced nature and unreliability of the sole inculpatory evidence is credited, and (c) violating Brady by failing to disclose threats and inducements used to shape witness testimony—even when the witnesses were known to the defense.

Plaintiffs George Clark and Kevin Harrington served nearly two decades in prison for murder convictions later vacated after the Wayne County Prosecution Integrity Unit concluded they “did not receive a fair trial” because the case turned on a single eyewitness who claimed police coercion, alongside alleged coercion of other witnesses. The defendants—Detective Anthony Abdallah and Sergeant Kevin Smith—appealed the district court’s denial of qualified immunity on claims of fabrication of evidence, federal and Michigan malicious prosecution, and (as to Abdallah) Brady violations.

Key issues

  • Appellate jurisdiction: What parts of a qualified-immunity denial are reviewable on interlocutory appeal, and whether evidentiary admissibility disputes are reviewable.
  • Fabrication (Fourteenth Amendment): Whether coerced witness narratives can amount to “knowing fabrication” and whether prosecutorial use of statements breaks causation.
  • Malicious prosecution (Fourth Amendment): Whether probable cause can rest on testimony claimed to be fabricated or so unreliable that officers had “apparent reason” to doubt it.
  • Brady (Fourteenth Amendment): Whether threats/inducements used to obtain witness cooperation are suppressible impeachment material, and whether “essential facts” were available to defense counsel.

II. Summary of the Opinion

The Sixth Circuit affirmed the denial of qualified immunity and remanded for trial on the remaining claims.

  • Fabrication: The court held plaintiffs presented triable issues that defendants knowingly fabricated Stewart’s and Wiseman’s inculpatory testimony; a stand-alone Fourteenth Amendment fabrication claim is viable “without regard to probable cause.”
  • Federal malicious prosecution: Because the sole probable-cause evidence was Stewart’s testimony, and plaintiffs raised triable issues of fabrication/unreliability, probable cause could be found lacking.
  • Michigan malicious prosecution: The same probable-cause disputes precluded Michigan governmental immunity because “good faith” is incompatible with knowingly proceeding on fabricated/unreliable evidence.
  • Brady: The court held Abdallah could be liable for suppressing impeachment evidence: undisclosed threats/inducements to Wiseman and coercive tactics toward Moore, rejecting the argument that witness availability alone defeats suppression.
  • Jurisdiction limitation: The panel lacked jurisdiction to review the district court’s hearsay/admissibility ruling regarding Stewart’s trial testimony.

III. Analysis

A. Precedents Cited (and how they shaped the result)

1) Qualified immunity framework

The court applied the orthodox two-step inquiry: constitutional violation and clearly established law. It cited Harlow v. Fitzgerald for the core standard (immunity unless violation of clearly established law), and relied on Tanner v. Walters to place the burden on plaintiffs at summary judgment and to distinguish Fourth Amendment malicious-prosecution claims from stand-alone due-process fabrication claims. Ashcroft v. al-Kidd reinforced the need for clearly established law, while later clearly-established analysis was anchored by the due-process line of cases discussed below.

2) Interlocutory appellate jurisdiction in qualified-immunity denials

The panel’s jurisdiction discussion is driven by the Supreme Court’s collateral-order doctrine line: Johnson v. Jones (no interlocutory review of “evidence sufficiency” determinations), Mitchell v. Forsyth (qualified immunity can be immediately appealable when presenting a purely legal issue), and Plumhoff v. Rickard (reviewable legal questions include whether the facts as assumed show a constitutional violation and whether the right was clearly established).

Sixth Circuit applications—Gillispie v. Miami Township, Adams v. Blount County, Diluzio v. Village of Yorkville, United Pet Supply, Inc. v. City of Chattanooga, Ashford v. Univ. of Michigan, and Heeter v. Bowers—were used to police the boundary between reviewable legal questions and unreviewable factual disputes. The “blatantly contradicted” exception from Scott v. Harris, as channeled through Adams v. Blount County, was identified but found inapplicable.

Critically, the court relied on Ellis v. Washington County (and consistent unpublished/cited examples Harmon v. Hamilton County and Bertl v. City of Westland) to hold that when an interlocutory qualified-immunity appeal “rests crucially” on evidentiary admissibility (here, hearsay), the court lacks jurisdiction to review that ruling. This preserved the district court’s evidentiary posture for trial and constrained defendants to legal arguments that assumed admissibility.

3) Fabrication of evidence—Fourth vs. Fourteenth Amendment

The court reaffirmed that fabrication theories can arise under both amendments, citing Hoskins v. York. For Fourth Amendment seizure, it cited Manuel v. City of Joliet (detention without probable cause violates the Fourth Amendment even after legal process) and King v. Harwood (probable cause is undermined when based on fabricated evidence presented to a grand jury/judge). But for the stand-alone fair-trial due-process claim, the panel leaned on Jackson v. City of Cleveland (quoting Gregory v. City of Louisville) for the test: knowing fabrication plus reasonable likelihood of affecting the jury.

The opinion also used out-of-circuit authorities to illustrate why coercion and investigative pressure can support an inference of knowing falsity: Fields v. Wharrie and Devereaux v. Abbey were cited to show coercion is not “legally irrelevant” and can be a tool to induce fabrication. The Eighth Circuit’s Winslow v. Smith supported liability where officers pressure witnesses to fill investigative gaps or ignore innocence evidence. The court also cited United States v. Burch for the general point that inconsistent statements go to credibility for the factfinder, not legal impossibility. Stewart v. Wolfenbarger was used by analogy to highlight the significance of alibi evidence in “thin” cases.

Defendants’ principal Sixth Circuit comparator, Price v. Montgomery County, was distinguished on factual grounds: in this case, officers allegedly knew of an alibi witness, witnesses were especially vulnerable (withdrawal/addiction), and the record supported a broader “pattern” of coercive conduct. The panel treated defendants’ remaining arguments as (unreviewable) evidence-sufficiency disputes under Johnson v. Jones.

4) Causation and the prosecutor’s role

Defendants argued prosecutorial use of Stewart’s interview materials was a superseding cause. The court rejected this based on its own binding precedent: Jackson v. City of Cleveland and Monson v. City of Detroit (causation is satisfied where officers procure a false statement that later compels conforming testimony and is introduced at trial). McDonough v. Smith was treated as consistent (plaintiff must show proceedings and liberty deprivations were caused by fabrication).

The opinion also addressed and discounted out-of-circuit causation authorities invoked by defendants: Wray v. City of New York and Townes v. City of New York (distinguishable evidentiary contexts), and Evans v. Chalmers (different standard and tension with Sixth Circuit’s Sykes v. Anderson, which allows liability when falsehoods influence continued detention).

5) Clearly established law for fabrication

The panel applied the “not too general” admonition from Ashcroft v. al-Kidd, as operationalized through Guertin v. Michigan (quoting Mullenix v. Luna). It nevertheless found the right clearly established by longstanding Supreme Court due-process doctrine: Pyle v. Kansas (police threats causing perjury deprive defendants of constitutional rights), Mooney v. Holohan (conviction through known perjured testimony violates due process), and Miller v. Pate (knowing presentation of false evidence violates due process even without “planting”). The court framed the operative “particular conduct” as knowingly procuring/using false testimony to secure conviction, not merely “using hard interrogation tactics.”

6) Fourth Amendment malicious prosecution—probable cause

The malicious-prosecution elements were taken from Tanner v. Walters (quoting France v. Lucas). For probable cause, the court invoked Beck v. Ohio (reasonably trustworthy information standard), and the Sixth Circuit eyewitness reliance rule from Ahlers v. Schebil (officers may rely on eyewitness identification unless there is apparent reason to doubt truth/accuracy). It emphasized that probable cause cannot rest on fabrication, consistent with Manuel v. City of Joliet and King v. Harwood. The “jury question” nature of probable cause was supported by Ouza v. City of Dearborn Heights (quoting Gardenhire v. Schubert).

7) Michigan malicious prosecution and Michigan governmental immunity

The Michigan malicious-prosecution elements came from Moldowan v. City of Warren (quoting Payton v. City of Detroit). Michigan governmental immunity for intentional torts was analyzed under Odom v. Wayne County: immunity requires discretionary action within scope and in good faith; malice/improper purpose defeats immunity. Latits v. Phillips illustrated that honest mistaken probable cause can still be good faith—distinguished here because plaintiffs’ theory was knowing fabrication/unreliability.

8) Brady obligations, “suppression,” and witness availability

The basic Brady framework came from Brady v. Maryland, with elements and materiality drawn via Jackson v. City of Cleveland (quoting Strickler v. Greene). Police officers’ Brady duties were grounded in Moldowan v. City of Warren.

For the “essential facts”/availability doctrine, the court cited: McNeill v. Bagley (exclusive control concept), Coe v. Bell (quoting United States v. Clark) (no Brady violation where defendant knew/should have known essential facts or evidence available from another source), Owens v. Guida (no suppression where defendant knew underlying facts), and Barton v. Warden (Brady is not satisfied by giving a “trail of crumbs”; defense must be able to find the “cookie”).

Rejecting a broad rule that defense witness availability defeats Brady, the court cited: Boss v. Pierce (rejecting the “untenable” view that any information possessed by a defense witness is necessarily “available” for Brady), and Johnson v. Folino (availability is not ensured merely because a witness can be cross-examined). It relied on Barton v. Warden to show suppression can exist even when a witness is identifiable/locatable, if the withheld “essential facts” are not known.

On the substance of withheld impeachment evidence, the panel endorsed (and applied) the concept that threats/inducements and coercive tactics are classic impeachment material: Anderson v. City of Rockford (Brady requires disclosure of coercive tactics used to obtain a statement), Avery v. City of Milwaukee (such evidence can help prove statements were false and undermine core witnesses), and Sixth Circuit prejudice analysis from Harris v. Lafler (prejudice where the case “hinges on the testimony of one witness,” and defense counsel cannot prove suspected promises/threats). It also noted a discoverability nuance supported by Jalowiec v. Bradshaw and Crivens v. Roth (witness fear/retaliation may prevent defense discovery, supporting “suppression”).

Responding to the dissent’s attempt to classify allegations as “not evidence,” the majority cited: Bell v. Bell (en banc) (Brady covers informal or tacit agreements), United States v. Rodriguez (Brady obligations exist regardless of whether information is recorded in tangible form), Jackson v. City of Cleveland (Brady claim may rest on allegations that a false statement “had been coerced”), and Harris v. Lafler (informal threats/promises must be disclosed). The dissent’s reliance on Coleman v. Bradshaw was distinguished as a habeas credibility/standard-of-review context; the majority also cited Berryman v. Rieger to emphasize that credibility disputes are not for interlocutory review. Finally, it invoked Heck v. Humphrey to answer “floodgates” concerns by noting the favorable-termination barrier to many Brady-based § 1983 actions.

9) The dissent’s Brady framing (and the majority’s response)

Judge Batchelder agreed on fabrication and malicious prosecution but would have reversed on Brady, reasoning (i) Moore’s and Wiseman’s accounts were not the kind of “evidence” Brady contemplates, and (ii) even if they were, the information was “available” because both witnesses were defense witnesses or subject to cross-examination. The dissent drew on Strickler v. Greene, United States v. Agurs, Arizona v. Youngblood, Henness v. Bagley, Coleman v. Mitchell, and Coleman v. Bradshaw for limiting principles. The majority answered that (a) Brady is not restricted to documentary items and can cover unwritten coercion/benefits, (b) “availability” turns on whether the defense knew the “essential facts” enabling effective use, and (c) credibility contests over coercion allegations are for the jury at trial, not the appellate court in an interlocutory posture.

B. Legal Reasoning

1) Jurisdiction first: separating “law” from “fact” and excluding admissibility disputes

The opinion is structured around a threshold constraint: the Sixth Circuit can review only purely legal questions in an interlocutory qualified-immunity appeal. It therefore:

  • Accepted the district court’s fact view where defendants’ counter-narratives would require reweighing evidence (barred by Johnson v. Jones).
  • Declined to review the district court’s hearsay ruling admitting Stewart’s prior testimony because admissibility disputes are not reviewable on this interlocutory track (Ellis v. Washington County).

This “jurisdictional discipline” matters doctrinally: it forces the qualified-immunity analysis to proceed on the assumed evidentiary record as it will be presented to the jury, rather than on appellate relitigation of the record’s reliability.

2) Fabrication: coercion plus contextual indicators can support an inference of “knowing falsity”

The court agreed with the district court that “coercion alone” may not always establish fabrication, but held the record contained “more than just coercion”: repeated denials, shifting stories, an unrecorded break, knowledge of an alibi witness, witnesses’ vulnerability (withdrawal/addiction), and alleged threats involving children and jail. These combined circumstances could allow a jury to infer the officers knew (or had reason to know) the inculpatory story was false or highly unreliable and nevertheless caused its use at trial.

The court’s analogy is telling: just as multiple symptoms can support a diagnosis where each alone might not, multiple indicia of unreliability and coercive pressure can support an inference of knowing fabrication even if each factor alone is not dispositive.

3) Causation: prosecutorial use of the statement does not automatically sever officer liability

Defendants argued that because the prosecutor possessed recordings of Stewart’s interview and introduced them, the prosecutor’s actions were a superseding cause. The court rejected this as foreclosed by Jackson v. City of Cleveland and Monson v. City of Detroit: once officers procure a false account that shapes sworn testimony and is introduced at trial, the causal chain can remain intact. The mere fact that the prosecutor had access to the interview does not, as a matter of law, cleanse the taint if the testimony itself was knowingly fabricated by coercive procurement.

4) Malicious prosecution: when probable cause depends on a single contested witness, fabrication/unreliability disputes are central

Because Stewart’s testimony was the “sole, substantive evidence” at preliminary examination and effectively carried through trial by admission of prior testimony, the probable-cause question could not be resolved as a matter of law. If the testimony was fabricated, probable cause fails; even if not fabricated, the officers may have had “apparent reason” to doubt its reliability (Ahlers v. Schebil). Thus, summary judgment and qualified immunity were properly denied.

5) Brady: threats and inducements are impeachment material, and “availability” requires knowledge of essential facts

The court treated the undisclosed coercion/benefits alleged as classic impeachment evidence that could (i) impeach Wiseman, (ii) impeach Abdallah, and (iii) cast doubt on Stewart and the integrity of the investigation—especially given the “hinge” nature of Stewart’s account in a case with no physical evidence.

On suppression, the court drew a key line: a witness’s mere availability for interview or cross-examination is not equivalent to the defense knowing the essential facts of threats, promises, bond manipulation, or coercive detention. The first trial testimony, in the court’s view, did not reveal the coercion; indeed, Wiseman denied being threatened or promised anything, paralleling the dynamic in Harris v. Lafler, where counsel suspected but could not prove undisclosed inducements.

C. Impact

1) Reinforcement of stand-alone due-process fabrication claims

The decision strengthens the practical viability of Fourteenth Amendment fabrication claims in the Sixth Circuit by reiterating that they are not tethered to probable cause. This matters in wrongful-conviction litigation where probable cause may have existed at some early stage, yet the trial’s fairness is alleged to have been corrupted by knowingly false testimony.

2) Causation doctrine: prosecutorial awareness is not an automatic shield

By rejecting a categorical “superseding cause” defense based on the prosecutor’s possession or use of interview materials, the opinion signals that officers cannot rely on prosecutorial decisions to immunize the procurement of false testimony—especially when the alleged fabrication operates through the witness’s compelled conformity and subsequent use at trial.

3) Brady in practice: disclosure duties reach coercion/inducements and are not defeated by witness availability

The ruling’s most consequential aspect may be its application of Brady to undisclosed coercive tactics and inducements involving witnesses connected to the defense. Prosecutors and police in the Sixth Circuit should treat threats, child-custody leverage, promises of release, bond assistance, or intimidation used to shape testimony as Brady/Giglio-type impeachment material requiring disclosure, even if defense counsel knows the witness and can subpoena them—because “essential facts” may still be unknown.

4) Interlocutory appeals: evidentiary admissibility is largely insulated at the qualified-immunity stage

The jurisdiction holding underscores that defendants cannot typically obtain interlocutory review by reframing hearsay/admissibility objections as qualified-immunity questions. That places increased strategic weight on district-court evidentiary rulings and on building a record for post-trial review rather than collateral-order appeals.

IV. Complex Concepts Simplified

  • Qualified immunity: A shield from damages (and often from trial) unless plaintiffs show a constitutional violation of clearly established law.
  • Interlocutory appeal: An appeal before final judgment; allowed for qualified immunity only for “purely legal” issues—not disputes about whose facts are believable.
  • Fabrication of evidence (Due Process): Knowingly creating/using false evidence where it likely influenced the jury’s decision; focused on trial fairness rather than probable cause.
  • Malicious prosecution (Fourth Amendment): Wrongfully subjecting a person to legal proceedings without probable cause, causing liberty deprivation after the initial arrest.
  • Brady material: Favorable evidence (including impeachment) that the State must disclose; suppression can exist even if a witness is known, if the defense lacks the “essential facts” to uncover and use the information effectively.
  • Superseding cause: A later actor’s decision that breaks causation; this opinion rejects a broad version of that concept where officers allegedly procured false testimony later used at trial.

V. Conclusion

The Sixth Circuit’s decision in Clark v. Abdallah is a procedurally disciplined, jury-centered opinion: it confines interlocutory qualified-immunity review to legal questions, refuses to revisit evidentiary admissibility at that stage, and holds that a record of alleged witness threats, inducements, and contextual unreliability can support trial-worthy claims for (1) due-process fabrication, (2) Fourth Amendment malicious prosecution where probable cause hinges on the contested testimony, and (3) Brady violations for undisclosed coercion/benefits—even when the affected witnesses were known to the defense.

In practical terms, the opinion warns investigators and prosecutors that coercive witness handling can generate overlapping constitutional exposure: it can create fabricated evidence, destroy probable cause, and constitute suppressible impeachment material. The case now returns for a merits trial where a jury—rather than the appellate court—will resolve credibility, coercion, and knowledge.