Armendariz v. City of Colorado Springs: Digital-Search Warrants Must Supply a Crime-Specific Nexus and Seizure-Limiting Principles (Including for Keyword and Facebook Data)

Introduction

In Armendariz v. City of Colorado Springs (10th Cir. Mar. 12, 2026) (revised opinion issued after panel rehearing), the Tenth Circuit addressed the Fourth Amendment’s particularity requirement in the modern setting of (i) warrants to seize multiple electronic devices from a home, (ii) warrants authorizing forensic searches of those devices using broad keyword and file categories, and (iii) warrants compelling production of broad categories of Facebook account content.

The plaintiffs were Jacqueline Armendariz (a march participant) and the Chinook Center (a nonprofit associated with local activism). They sued the City of Colorado Springs and several officers under 42 U.S.C. § 1983, alleging that three warrants were overbroad and thus unconstitutional. Armendariz also pursued an injunctive claim against the FBI for retaining copied data. The Chinook Center additionally brought a Stored Communications Act claim (18 U.S.C. §§ 2701-13) and state-law claims. After the district court dismissed the entire case on Rule 12(b)(6) (including on qualified immunity), plaintiffs appealed.

The core issues were: (1) whether the warrants were plausibly overbroad under the Fourth Amendment; (2) whether any violations were clearly established such that qualified immunity could not be granted at the pleading stage; (3) whether municipal liability claims against the City could proceed; and (4) whether non-constitutional claims (SCA and state-law claims) were properly dismissed.

Summary of the Opinion

  • Qualified immunity reversed for all four officers involved in the three warrants. The court held that plaintiffs plausibly alleged the warrants were overbroad and that the right was clearly established.
  • Armendariz’s first warrant (home search/seizure of devices) held plausibly overbroad. The affidavit failed to establish even arguable probable cause (a crime-specific nexus) to seize the broad set of devices based largely on speculative “training and experience” generalities.
  • Armendariz’s second warrant (forensic search of device data) held plausibly overbroad. The keyword list and file categories, as drafted, lacked sufficient limiting principles (crime nexus, meaningful temporal limits, and a search/seizure distinction).
  • Chinook Center Facebook warrant held plausibly overbroad. Time-limiting alone did not cure a demand for “all” posts/chats/events without tying the seizure to evidence of the specified crimes; an incorporated affidavit could not “save” the warrant where it contradicted the warrant’s unambiguous breadth.
  • Municipal liability claims revived. Because the district court dismissed City claims only due to the absence of an underlying constitutional violation, the Tenth Circuit reversed and remanded for the district court to consider Monell issues in the first instance.
  • FBI injunctive claim affirmed on waiver. Armendariz failed to properly challenge the district court’s Rule 41(g) rationale on appeal.
  • FTCA-related state claim against the United States affirmed on waiver. Armendariz did not appeal the exhaustion-based dismissal rationale.
  • SCA claim and remaining state claims reinstated. The district court’s dismissal rested on its Fourth Amendment analysis and supplemental-jurisdiction logic, both undermined by the Tenth Circuit’s constitutional rulings.

Analysis

Precedents Cited

1) Pleading-stage posture and documents considered

  • Johnson v. Smith, 104 F.4th 153 (10th Cir. 2024): reaffirmed that, on a motion to dismiss, well-pleaded allegations are accepted as true and viewed favorably to the nonmovant.
  • Gee v. Pacheco, 627 F.3d 1178 (10th Cir. 2010): allowed consideration of warrant materials attached to motions to dismiss when referenced in the complaint and central to claims.
  • Stark v. Reliance Standard Life Ins., 142 F.4th 1252 (10th Cir. 2025); Truman v. Orem City, 1 F.4th 1227 (10th Cir. 2021): set the de novo dismissal standard and the requirement to credit well-pleaded facts.

2) Qualified immunity framework for warrants

  • Allstate Sweeping, LLC v. Black, 706 F.3d 1261 (10th Cir. 2013): described qualified immunity generally.
  • Thomas v. Kaven, 765 F.3d 1183 (10th Cir. 2014): noted that qualified immunity can be resolved at the 12(b)(6) stage but defendants face a “more challenging” posture than at summary judgment.
  • Brown v. City of Tulsa, 124 F.4th 1251 (10th Cir. 2025); Est. of Taylor v. Salt Lake City, 16 F.4th 744 (10th Cir. 2021): reiterated the two-prong test (plausible violation + clearly established law).
  • Stonecipher v. Valles, 759 F.3d 1134 (10th Cir. 2014) and Messerschmidt v. Millender, 565 U.S. 535 (2012): emphasized that a magistrate’s issuance is a strong indicator of objective reasonableness, but not an absolute shield.
  • United States v. Leon, 468 U.S. 897 (1984): supplied the four recognized circumstances in which reliance on a warrant is unreasonable (including facial deficiency in particularity).

3) Particularity/overbreadth and nexus doctrine (Tenth Circuit core)

  • Mink v. Knox, 613 F.3d 995 (10th Cir. 2010): central to the court’s “overbroad if no nexus between crime and items seized” formulation; also used to anchor clearly established law.
  • Voss v. Bergsgaard, 774 F.2d 402 (10th Cir. 1985): reinforced that a warrant must confine scope to evidence of a specific crime with demonstrated probable cause.
  • United States v. Cotto, 995 F.3d 786 (10th Cir. 2021): clarified that overbreadth authorizes seizures without probable cause.
  • United States v. Biglow, 562 F.3d 1272 (10th Cir. 2009): explained nexus is fact-dependent and analyzed by reasonable inferences; used as a baseline, not a license for speculation.
  • United States v. Leary, 846 F.2d 592 (10th Cir. 1988): repeatedly invoked for the rule that if specificity is available it must be used, and warrants must let officers distinguish seizable from non-seizable items.
  • Cassady v. Goering, 567 F.3d 628 (10th Cir. 2009): used for the proposition that obvious overbreadth makes clearly-established analysis “easily satisfied.”

4) Digital-search-specific authorities (within and beyond the circuit)

  • United States v. Otero, 563 F.3d 1127 (10th Cir. 2009): key statement that computer warrants must “affirmatively limit” the search to evidence of specific crimes or types of material.
  • United States v. Palms, 21 F.4th 689 (10th Cir. 2021): emphasized the need for a “limiting principle” and warned against wide-ranging exploratory digital searches.
  • United States v. Russian, 848 F.3d 1239 (10th Cir. 2017) and United States v. Wagner, 951 F.3d 1232 (10th Cir. 2020): recognized that the Fourth Amendment does not generally require ex ante search protocols; the court distinguished “search strategy” from “items authorized to be seized.”
  • United States v. Suggs, 998 F.3d 1125 (10th Cir. 2021); United States v. Sadlowski, 948 F.3d 1200 (10th Cir. 2020): addressed incorporation requirements and how incorporated affidavits can (sometimes) cure ambiguity by supplying limiting language.
  • United States v. Mora, 989 F.3d 794 (10th Cir. 2021): particularly influential; condemned probable-cause showings that rely on general “experience” statements untethered to the defendant’s circumstances.
  • United States v. Griffith, 867 F.3d 1265 (D.C. Cir. 2017): persuasive authority for rejecting “blanket authorization” to seize electronic devices absent a concrete showing those items likely contain crime evidence.

5) Home-search primacy and privacy interests

  • Kyllo v. United States, 533 U.S. 27 (2001) (quoting Silverman v. United States, 365 U.S. 505 (1961)): reinforced that the home is at the core of Fourth Amendment protection.

6) Location data sensitivity

  • Carpenter v. United States, 585 U.S. 296 (2018): quoted for the proposition that location data reveals an “intimate window” into life and associations, supporting the court’s skepticism of broad location-data seizures.
  • United States v. Barajas, 710 F.3d 1102 (10th Cir. 2013): used to doubt broad GPS/location requests absent adequate justification.

7) Social media (Facebook) warrant scrutiny

  • United States v. Blake, 868 F.3d 960 (11th Cir. 2017) and United States v. Shipp, 392 F. Supp. 3d 300 (E.D.N.Y. 2019): cited for heightened concern about Facebook’s breadth and its categorization/searchability (reducing any claim that “everything” must be seized to find evidence).
  • United States v. Purcell, 967 F.3d 159 (2d Cir. 2020): distinguished because it upheld breadth where illegality “pervaded” the entire account—something not shown here.
  • United States v. Ulbricht, 858 F.3d 71 (2d Cir. 2017): cited for limiting seizure to “evidence, fruits, and instrumentalities” of specified offenses.
  • United States v. Turner, 2022 WL 195083 (D. Nev. Jan. 21, 2022) and United States v. Liburd, 2018 WL 2709199 (E.D.N.Y. June 5, 2018): distinguished because those warrants/affidavits supplied objective standards to segregate responsive from non-responsive material.
  • United States v. Beaumont, 972 F.2d 553 (5th Cir. 1992): used for the proposition that affidavits can clarify ambiguity but cannot rewrite a warrant’s unambiguous breadth.

8) Waiver on appeal

  • United States v. Martinez, 92 F.4th 1213 (10th Cir. 2024) and Nixon v. City & Cnty. of Denver, 784 F.3d 1364 (10th Cir. 2015): applied to hold Armendariz waived challenges by not confronting the district court’s reasoning.

Legal Reasoning

1) The court’s unifying move: apply traditional particularity principles with digital-specific “limiting principle” rigor

The opinion is best read as synthesizing two lines of doctrine: (1) the “nexus to a specific crime” requirement from cases like Mink v. Knox, Voss v. Bergsgaard, and United States v. Leary, and (2) the digital-search insistence—drawn from United States v. Otero and United States v. Palms—that warrants must include affirmative limits that prevent general rummaging through the “enormous amounts” of personal data stored digitally.

2) First Armendariz warrant: no arguable nexus to seize “all the devices” from the home

The first warrant authorized seizure of a wide set of devices (phones, computers, tablets, thumb drives, and external hard drives) from the home. The court accepted that the affidavit showed Armendariz had an online presence and likely used at least some device(s). But it found the step from “social media activity exists” to “all manner of storage devices in the home likely contain evidence of a spontaneous assault at a march” was speculative—especially where the affidavit leaned on generalized statements about how “people” back up phones to computers or store data on various media.

Crucially, the court treated United States v. Mora as controlling in spirit: experience-based generalities untethered to the suspect’s crime and circumstances do not supply the constitutionally required nexus. The panel concluded that (even with magistrate approval) the warrant was so deficient in particularizing the things to be seized, in relation to the crime, that officers could not reasonably presume validity under the Leon framework as imported into qualified-immunity analysis.

3) Second Armendariz warrant: keyword and file categories lacked meaningful limits—especially as to “seizure”

The second warrant’s design featured two mechanisms:

  • A file-category search (photos, videos, messages, emails, and location data) for a two-month period; and
  • A keyword search for broad and politically laden terms and many proper names, explicitly “regardless of the time period.”

The opinion identifies three related failures.

  1. Crime linkage: Terms like “protest,” “right,” “police,” or names of organizers could sweep vast quantities of irrelevant data, including purely expressive or associational material, without being tied to evidence of the attempted assault.
  2. Failure to use available narrowing information: Borrowing from United States v. Leary, the court treated the warrant’s breadth as a constitutional choice: the officers had the ability to narrow by event, target, or time, yet drafted language that maximized discretion.
  3. Search vs. seizure collapse: The court emphasized that the warrant effectively authorized seizure of everything responsive to the keyword search, rather than authorizing a search method whose results would then be filtered to seizable evidence. This collapse was bolstered by the court’s reliance on United States v. Zelaya-Veliz (discussing the two-step “search broadly / seize narrowly” logic for digital data).

The panel also rejected the argument that an incorporated affidavit cured the defects, emphasizing that a curing affidavit must actually introduce a limiting principle and cannot be treated as silently rewriting a warrant whose operative attachments authorize broad seizure.

4) Facebook warrant: time restriction alone was not enough; “all content” plus affidavit contradiction was fatal

The Facebook warrant did something common in practice but newly disfavored here: it limited the time window (seven days) but sought “all” posts, “all” Messenger chats, and “all” events in that window. The court treated Facebook’s architecture as relevant to Fourth Amendment reasonableness: because Facebook data is categorized and searchable (and not subject to the same “hide evidence in obscure folders” concerns of device imaging), a demand for “all” content is harder to justify.

Two doctrinal points are especially consequential:

  • Affidavit incorporation has limits: The court held that an affidavit cannot “save” a warrant when it directly contradicts the warrant’s unambiguous breadth (here, “all” content). This builds on incorporation cases like United States v. Suggs and limiting principles discussed in United States v. Sadlowski, while invoking United States v. Beaumont for the “clarify ambiguity vs. change meaning” distinction.
  • Clearly established law does not require “Facebook-specific” precedent: The court relied on Mink v. Knox, Leary, Voss, and Cassady to conclude that the general particularity rule applies “with obvious clarity” to a warrant that authorizes seizure of plainly irrelevant materials.

5) The dissent’s technological-fair-notice critique

Judge Bacharach agreed that the Facebook warrant lacked arguable nexus but would have granted qualified immunity for the first two warrants. The dissent’s theme was that Fourth Amendment doctrine has not “kept up” with digital realities, so officers could reasonably rely on judicial approvals in an underdeveloped area. The majority’s response was effectively that the deficiency was not a subtle “digital” problem but a classic particularity/nexus failure: broad device seizure and broad data seizure, justified by generalized experience and politically oriented keywords, without practical limits tethered to the crime.

Impact

1) Warrant drafting in the Tenth Circuit: concrete constraints are no longer optional

Armendariz will likely push agencies and prosecutors to implement more disciplined digital-warrant templates, including:

  • Device-seizure warrants: fact-based explanations for why each type/class of device is likely present and likely contains evidence of the particular offense (not merely that “people” store data on devices).
  • Forensic device-search warrants: explicit, crime-tethered limitations on what may be seized (not just searched), with careful attention to temporal scope and subject matter.
  • Keyword warrants: keywords must function as constraints; broad terms (“right,” “protest,” generic names) and “regardless of time period” phrasing risk being treated as facial overbreadth.
  • Social media warrants: “all posts/chats/events” is highly vulnerable even with a short time window; requests should be narrowed to communications with identified targets, categories that constitute “evidence/fruits/instrumentalities,” or objective criteria for responsiveness.

2) Litigation consequences: easier pleading-stage survival against qualified immunity for facially broad digital warrants

The decision is notable procedurally: it is a Rule 12(b)(6) reversal that denies qualified immunity based on the face of warrant documents. Plaintiffs challenging digital warrants may rely on Armendariz for the proposition that certain breadth problems are “facial” and thus resolvable without discovery—cutting against the usual expectation that qualified immunity is decided on a more developed record.

3) Municipal liability exposure and policy reform pressure

Because the panel revived claims against the City and remanded, municipalities in the circuit face increased incentives to adopt (and document) narrower warrant practices, training, and supervisory review—especially where protests and First-Amendment-adjacent contexts could prompt allegations of retaliatory or ideological fishing expeditions.

4) Stored Communications Act enforcement may track Fourth Amendment rigor in practice

The district court treated the Stored Communications Act as no more demanding than the Fourth Amendment and dismissed on the same grounds. The reversal means SCA claims may rise or fall with warrant particularity in many cases—particularly those involving compelled social-media disclosures under 18 U.S.C. § 2703.

Complex Concepts Simplified

  • Particularity: The warrant must be specific enough that officers know exactly what they may search and seize—so the warrant does not become a “general warrant.”
  • Overbreadth (Fourth Amendment): A warrant is overbroad when it authorizes seizure of items for which there is no probable cause—often because it is not tied tightly to the crime being investigated.
  • Nexus: The required link between (a) the suspected offense and (b) the place/items to be searched/seized. “Someone involved in protests uses phones” is not, by itself, a nexus to seize all digital devices in a home.
  • Qualified immunity: Even if a warrant is unconstitutional, officers are liable only if the unconstitutionality was clearly established—i.e., obvious under existing precedent or clearly governed by it.
  • Arguable probable cause: In the qualified-immunity context, the question becomes whether reasonable officers could think probable cause existed, even if a court later disagrees.
  • Incorporated affidavit: A warrant can rely on an attached affidavit to supply details—but only if properly incorporated and only if the affidavit actually narrows/clarifies rather than contradicts an unambiguous command like “all messages.”
  • Search vs. seizure (digital): Officers may need to search broadly to find evidence, but they may seize/copy only what qualifies as evidence of the specified crime. A warrant that authorizes seizure of every “hit” from broad keywords risks becoming a privacy-invasive dragnet.
  • Waiver on appeal: If an appellant does not meaningfully address the district court’s stated reason for dismissal, the appellate court may affirm without reaching the merits.

Conclusion

Armendariz v. City of Colorado Springs is a significant Tenth Circuit statement that digital-era warrants remain bound by familiar Fourth Amendment constraints: there must be a concrete, crime-specific nexus and a genuine limiting principle—especially where warrants seek broad categories of digital devices, sweeping keyword-driven data, or “all” social-media content. Time windows alone may not cure breadth, and affidavits cannot salvage a warrant whose plain text authorizes indiscriminate seizure. The decision’s practical effect is to raise drafting standards for digital and social-media warrants and to make it harder to defeat overbreadth challenges at the pleading stage through qualified immunity alone.