Hash-Scan Private Search Doctrine: Government Viewing of a PhotoDNA-Flagged CSAM File Does Not Exceed the Private Search

1. Introduction

State v. Gasper addresses a modern Fourth Amendment problem: how the “private search doctrine” applies when a private platform detects suspected child sexual abuse material (CSAM) not by a human review, but by automated hash-based scanning. Snapchat flagged a single 16-second video uploaded from Gasper’s Snapchat account using Microsoft’s PhotoDNA, then sent a CyberTip to the National Center for Missing and Exploited Children (“NCMEC”), which forwarded the report to Wisconsin authorities. No Snapchat or NCMEC personnel viewed the video, but a DOJ analyst and later a county detective opened and watched it without a warrant before seeking a warrant to search Gasper’s home and devices.

After execution of a warrant, police found additional CSAM on Gasper’s cell phone; Gasper also admitted storing CSAM. The circuit court suppressed all CSAM evidence, reasoning (i) Riley v. California and Carpenter v. United States demanded a warrant, and (ii) hashing (framed as “MD5”) was unreliable due to collision risk. The court of appeals reversed on a different theory (terms-of-service and no reasonable expectation of privacy), but the Wisconsin Supreme Court ultimately affirmed the reversal on private-search-doctrine grounds and expressly vacated the court of appeals’ reasoning, removing its precedential value.

Key issues: (1) Whether Snapchat’s automated hashing constitutes a “private search” and (2) whether the government “exceeded the scope” of that private search by being the first humans to view the file without a warrant.

2. Summary of the Opinion

The Wisconsin Supreme Court holds that Snapchat performed a private search when it scanned and flagged the specific 16-second video as known CSAM. Applying United States v. Jacobsen, the court concludes law enforcement did not exceed the scope of Snapchat’s private search when it viewed the very same flagged file without a warrant, because any expectation of privacy in that file had already been frustrated and there was “virtual certainty” that viewing would reveal nothing of significance beyond what the private search had already identified.

Accordingly, the Fourth Amendment was not implicated, the suppression order was improper, and the case was remanded for further proceedings. The court did not reach: (i) whether Gasper had a reasonable expectation of privacy in the video or Snapchat account under Riley/Carpenter, (ii) whether terms of service affect Fourth Amendment expectations, or (iii) the good-faith exception to exclusion.

3. Analysis

3.1 Precedents Cited (and How They Shaped the Decision)

• United States v. Jacobsen, 466 U.S. 109 (1984)
  Role: The controlling framework. Jacobsen draws the line between private searches (not regulated by the Fourth Amendment) and later governmental review. The key inquiry is whether authorities obtained information “with respect to which the defendant’s expectation of privacy has not already been frustrated,” and whether government conduct exceeded the private search’s scope. Jacobsen also supplies the “virtual certainty” concept: government reinspection is not a Fourth Amendment search where it is virtually certain to reveal nothing beyond what the private search already exposed.
• Walter v. United States, 447 U.S. 649 (1980)
  Role: Used as a contrast case. Walter suppressed evidence where FBI projection of films was a “significant expansion” of a limited private intrusion because the private party could only infer the contents. The Wisconsin Supreme Court distinguished Walter by emphasizing that Snapchat’s PhotoDNA actually scanned and flagged the specific file, and that law enforcement viewed only that segregated file, making it virtually certain no additional significant information would be uncovered beyond the private scan’s revelation.
• Camara v. Mun. Ct. of City & Cnty. of San Francisco, 387 U.S. 523 (1967) and Carpenter v. United States, 585 U.S. 296 (2018)
  Role: Cited for the Fourth Amendment’s purpose as a limit on government power and overbroad surveillance. They inform the court’s framing but do not supply the operative test; Jacobsen does.
• Riley v. California, 573 U.S. 373 (2014)
  Role: A major premise in the circuit court’s suppression order (cell-phone privacy and warrant requirement). The Supreme Court sidestepped Riley by resolving the case via the private search doctrine (no Fourth Amendment search), thus making Riley’s warrant rule inapplicable on these facts.
• United States v. Miller, 982 F.3d 412 (6th Cir. 2020) and United States v. Reddick, 900 F.3d 636 (5th Cir. 2018)
  Role: Persuasive authorities endorsing application of Jacobsen to hash-matched CSAM files: where hashing identifies known CSAM with near certainty, police viewing does not exceed the private search. The Wisconsin Supreme Court aligns with these decisions.
• United States v. Wilson, 13 F.4th 961 (9th Cir. 2021) and United States v. Maher, 120 F.4th 297 (2d Cir. 2024)
  Role: The contrary federal appellate approach. Wisconsin rejects their rationale, characterizing it as inconsistent with Jacobsen because it treats law enforcement’s seeing “details” of CSAM as a constitutionally significant expansion even though the file has already been identified by private scanning as contraband with virtual certainty.
• United States v. Ackerman, 831 F.3d 1292 (10th Cir. 2016), United States v. Lichtenberger, 786 F.3d 478 (6th Cir. 2015), United States v. Rivera-Morales, 961 F.3d 1 (1st Cir. 2020), and United States v. Runyan, 275 F.3d 449 (5th Cir. 2001)
  Role: Used to articulate or elaborate the “virtual certainty” framing and scope-of-private-search analysis and to situate the case within broader private-search jurisprudence.
• State v. Payano-Roman, 2006 WI 47
  Role: Allocates the burden: once the State asserts a private search, the defendant must prove a government search occurred (preponderance standard). The court uses this to critique Gasper’s theory as failing to show the government exceeded Snapchat’s scan.
• Herring v. United States, 555 U.S. 135 (2009)
  Role: Quoted for exclusionary rule principles and culpability thresholds, although the majority declines to apply the good-faith analysis because it finds no constitutional violation.

3.2 Legal Reasoning

A. Re-centering the Fourth Amendment question on “scope” and “virtual certainty”

The court treats this as a Jacobsen case, not a Riley/Carpenter case. Once Snapchat’s scan is deemed a private search, the question becomes whether government viewing exceeded that private search. The majority’s pivotal move is to characterize PhotoDNA hashing as a genuine inspection of the file’s contents—performed by “Snapchat’s digital eyes”—such that human viewing of the segregated file adds nothing of Fourth Amendment relevance when it is virtually certain to confirm the same contraband status.

B. Rejecting the “no human saw it” theory

Gasper’s central contention is formal: because no Snapchat employee watched the video, a government employee’s first human viewing necessarily exceeds the private search. The majority rejects that as an incorrect premise: the private search doctrine does not require a human eyeball inspection by the private party. What matters is whether the private party already searched the item and whether the government learned anything of significance beyond what that private search revealed. Here, Snapchat’s process identified the file as known CSAM and turned over only that flagged file, making further discovery beyond the scan’s revelation virtually impossible.

C. Distinguishing Walter and limiting “expansion” concerns

Walter is treated as a case where the private search yielded only inferences. The Wisconsin Supreme Court emphasizes that Snapchat provided a single, specific file already flagged by a content-based scan, not a trove of unopened items. Thus, the government’s act is closer to confirming what was already disclosed than to expanding into unknown material.

D. Handling reliability and the circuit court’s MD5/collision findings

The circuit court’s suppression order rested partly on the claim that PhotoDNA used MD5 and that MD5 was categorically unreliable due to collision risk. The Supreme Court calls these factual findings “clearly erroneous,” emphasizing:

• the record indicated Snapchat used PhotoDNA, not MD5;
• the detective testified MD5 was not used to detect the file (though an MD5 value appeared in the CyberTip for investigative assistance);
• there was no record evidence of collision in this case or of PhotoDNA’s unreliability.

Notably, the majority also signals that private-search analysis is not necessarily controlled by whether the private party’s suspicion could be wrong; rather, it turns on what the private party actually did and whether government review stayed within that revealed scope.

E. Avoiding broader doctrines (expectation of privacy; terms of service; good faith)

The court expressly declines to decide whether Gasper had a reasonable expectation of privacy in the video or what effect Snapchat’s terms of service may have on Fourth Amendment analysis. That choice is doctrinally significant: it prevents the case from becoming a “third-party doctrine” or “contractual waiver” decision and instead pins the holding to Jacobsen’s private-search framework. Likewise, because the court finds no Fourth Amendment search, it does not reach the good-faith exception.

3.3 Impact

A. Wisconsin adopts the Miller/Reddick approach to hash-based CSAM tips

Wisconsin’s Supreme Court aligns itself with the line of authority treating hash-based identification of known CSAM as generating “virtual certainty” such that law enforcement’s warrantless viewing of the flagged file is not a Fourth Amendment search when confined to the same segregated file. This is the opinion’s core precedential effect: it operationalizes Jacobsen for automated, non-human private searches in the CSAM-hash context.

B. Limits and guardrails implied by the opinion

• File-limited scope: The reasoning repeatedly emphasizes the government viewed only “the one video provided” and not the broader Snapchat account or additional files. Future disputes will likely turn on whether police stayed confined to the exact file(s) flagged and provided.
• Technology-specific facts: The court relied on PhotoDNA’s function as described in the record and the detective’s experience that PhotoDNA matches had “never” been non-CSAM. Challenges to reliability may become more central in future cases where defendants build a record contesting scan accuracy.
• Vacatur of the court of appeals’ terms-of-service rationale: By vacating the court of appeals’ reasoning, the Supreme Court prevents a Wisconsin rule from emerging that terms of service can eliminate Fourth Amendment protections. That leaves the privacy/contract debate open in Wisconsin.

C. Practical effects on investigations

The opinion legitimizes a common investigative sequence: receive CyberTip with a flagged file, open and confirm the file, then seek broader warrants for devices and premises. The concurrence by Justice Ziegler underscores a policy concern: requiring a warrant to view the tip might prompt law enforcement to seek broader warrants immediately, increasing intrusion compared to a narrow confirmatory review.

D. Continued national disagreement

Because the opinion directly rejects United States v. Wilson and United States v. Maher, it deepens the doctrinal divide over whether viewing “details” of contraband is an “expansion” after a hash match. That split creates litigation risk when evidence crosses jurisdictions and may invite future U.S. Supreme Court review; in the meantime, Wisconsin litigants will face a state rule that is expressly on one side of the split.

4. Complex Concepts Simplified

• Private search doctrine: If a private party (not the government) searches something and then shows it to police, the Fourth Amendment generally does not require a warrant for police to look at the same thing—so long as police do not go beyond what the private party already searched.
• “Exceeding the scope” of the private search: Police “exceed the scope” when they learn something new that the private search did not already expose (e.g., opening additional containers, accessing more files, or exploring parts not previously searched).
• “Virtual certainty” (from United States v. Jacobsen): A common-sense judgment that the government will not find anything meaningful beyond what the private search already revealed—so police inspection is essentially confirmatory, not exploratory.
• Hash / PhotoDNA: A “hash” is like a fingerprint for a digital file. Traditional hashes (like MD5) change if the file changes even slightly. PhotoDNA is designed to identify known images or videos even if slightly altered, by analyzing features/pieces of the file and comparing them to a database of known CSAM.
• MD5 “collision”: A “collision” is when two different files produce the same hash value, potentially causing a false match. The circuit court treated collision risk as decisive; the Supreme Court rejected that factual premise on this record and emphasized PhotoDNA, not MD5, drove detection here.
• Reasonable expectation of privacy: The general test for whether a government act is a “search.” The majority avoids this question by holding that, under the private search doctrine, there was no Fourth Amendment search to begin with.

5. Conclusion

State v. Gasper establishes a clear Wisconsin rule for hash-based CSAM CyberTips: an automated PhotoDNA scan by a private platform can constitute a “private search,” and law enforcement may view the same segregated, flagged file without a warrant when there is “virtual certainty” it will reveal nothing of significance beyond what the private scan already disclosed. The opinion resolves the case through United States v. Jacobsen rather than through Riley/Carpenter cell-phone privacy principles or any terms-of-service theory (which the court of appeals had adopted but which the Supreme Court vacated).

Going forward, suppression litigation in Wisconsin will likely focus on boundaries the decision implicitly draws: whether police stayed confined to the exact flagged file(s), whether subsequent steps accessed broader accounts or devices without proper warrants, and whether defendants can meaningfully contest “virtual certainty” by building a record on scanning reliability or mismatch risk.