No Sender Privacy in Recipient-Stored Messages; Social-Media Warrants Must Be Narrowed by Time/Subject

Introduction

This direct appeal arose from Deshon Israel Bonnell’s conviction for first-degree premeditated murder for the shooting death of Joshua LaValley on the Mesabi Trail in Saint Louis County. The State’s case relied heavily on accomplice testimony (particularly Bailey French’s eyewitness account) and a large body of Facebook evidence—messages and account data obtained via a single warrant that targeted 12 Facebook accounts, including accounts owned by French, Anthony Howson, and two accounts purportedly owned by Bonnell (including “Pineapple Man”).

The appeal presented a mix of digital-search issues and trial issues. Most notably, the court confronted, as a matter of first impression, whether a person who sends an electronic message keeps a reasonable expectation of privacy in the copy stored in the recipient’s separate account or device. In addition, Bonnell challenged the warrant used to obtain data from his own Facebook accounts as a constitutionally invalid “general warrant” lacking temporal and subject-matter limits. He also challenged evidentiary rulings admitting (i) social-media evidence and (ii) “prior bad acts” evidence under Minnesota Rule of Evidence 404(b). Finally, he argued the evidence was insufficient because the State relied on accomplice testimony and circumstantial corroboration.

Summary of the Opinion

The court affirmed Bonnell’s conviction, but it did so after drawing several important constitutional lines:

• Recipient-account searches: Bonnell’s Fourth Amendment and state constitutional protections were not triggered by law enforcement searches of French’s and Howson’s Facebook accounts because Bonnell had no ownership interest in those accounts and, critically, because he had no reasonable expectation of privacy in the messages once received and stored in the recipients’ accounts.
• Bonnell’s own accounts: Bonnell did have a reasonable expectation of privacy in digital copies stored in his own two Facebook accounts. The warrant authorizing those searches was invalid for lack of any temporal or subject-matter limitations; thus the searches violated both the Fourth Amendment and article I, section 10.
• Harmless error: Despite constitutional error, the court held the admission of evidence obtained exclusively from Bonnell’s accounts was harmless beyond a reasonable doubt, largely because the incriminating messages also existed (lawfully) in French’s account and because the remaining evidence of guilt was substantial.
• Rule 404(b): The court assumed without deciding that admitting certain constitutionally obtained messages as “immediate episode” or “Spreigl” evidence was error, but held any error was harmless.
• Accomplice corroboration & sufficiency: The court held corroboration was “weighty enough” to restore confidence in French’s eyewitness testimony, and that the State’s evidence was sufficient to sustain the conviction.

Analysis

1) Precedents Cited

A. “Personal” nature of Fourth Amendment rights and expectation-of-privacy framework

The court grounded its analysis in the principle that Fourth Amendment rights are personal and require a legitimate expectation of privacy in the place searched:

• In re Welfare of B.R.K.: Cited for the requirement that an individual must personally have an expectation of privacy in the place searched and that the expectation be reasonable.
• State v. Perkins: Reinforced that Fourth Amendment protections are not triggered absent a legitimate expectation of privacy in the invaded space.
• State v. Carter: Framed the “standing” concept in Minnesota’s Fourth Amendment jurisprudence: without a legitimate expectation of privacy in the area searched, a defendant may not contest the search.

These cases supplied the doctrinal “gatekeeping” function for Bonnell’s attempt to suppress evidence taken from other people’s Facebook accounts—an argument the court rejected once it announced the new rule about sender privacy in recipient-stored messages.

B. Sender privacy after delivery: analogies to letters and adoption for electronic messages

To answer the first-impression question—whether a sender retains privacy in the recipient-stored copy—the court relied heavily on authorities treating delivered communications as no longer protected from the sender’s perspective:

• State v. Patino: The Rhode Island Supreme Court’s reasoning was central—once received, the sender cannot control the recipient’s use or disclosure, undermining any continued reasonable expectation of privacy in the recipient’s stored copy.
• Guest v. Leis and United States v. King: The Sixth Circuit’s analogy between delivered email and delivered letters (“expectation of privacy ordinarily terminates upon delivery”) was treated as persuasive and consistent with broader federal consensus.
• Federal “letter” line: United States v. Dunning, United States v. Gordon, United States v. Knoll, and Ray v. U.S. Dep't of Just. were cited to show uniform treatment of physical letters: once delivered, the sender’s privacy expectation ends.

The Minnesota Supreme Court imported this delivery-terminates-sender-privacy logic to modern messaging, resulting in a clear, administrable rule: once the message is received and stored in a separate account/device, the sender cannot claim a reasonable expectation of privacy in that stored copy.

C. Carpenter, dicta, and the boundaries of third-party doctrine arguments

Bonnell attempted to use Carpenter v. United States to argue for an exception when the government obtains the “modern-day equivalents” of papers and effects held by third parties. The court rejected the attempt on two precedent-based grounds:

• Sheehy Lee v. Kalis, City of Duluth v. Wendling, and Heller v. New York: These were used to emphasize that dicta (and a dissent) are not binding. Carpenter’s majority discussion of Justice Kennedy’s dissent-based “sensible exception” was treated as obiter dicta.
• Ex parte Jackson and United States v. Warshak: The court distinguished the “intermediary in transit” context (mail carrier; ISP-hosted emails) from the “end recipient” context. The court positioned the recipient as a separate, independent holder—not a conduit holding the sender’s effects.

This distinction is pivotal: the court’s new rule is expressly about recipient-stored copies and does not purport to resolve the privacy status of messages held by intermediaries while in transit or stored with service providers in ways that remain functionally within the sender’s (or user’s) private domain.

D. Warrants for digital accounts: particularity, temporal limits, and anti-general-warrant principles

Having held that Bonnell could not complain about searches of others’ accounts, the court turned to the search of Bonnell’s own accounts and relied on Minnesota’s evolving body of electronic-search warrant law:

• State v. Bradford: Provided the baseline that the particularity requirement prohibits general or exploratory searches.
• State v. Zielinski: Used for the proposition that if a warrant is not particular, the search is unconstitutional; the opinion emphasized the “invalid warrant” framing.
• State v. Sardina-Padilla: Functioned as the key Minnesota digital-search benchmark. The court contrasted the three-month temporal limitation upheld there with the absence of any temporal limit here, and reiterated Sardina-Padilla’s warning that authorizing “[a]ll content” standing alone would be “clearly impermissible.”
• State v. Miller and State v. Hannuksela: Supplied the flexible standard: particularity is evaluated case-by-case, considering the nature of the crime and whether a more precise description is possible; generic descriptions may be allowed only when as specific as circumstances permit.
• United States v. Blake: Served as an out-of-state analogue condemning Facebook warrants demanding virtually everything (messages, IP logs, photos, etc.) and suggesting obvious narrowing mechanisms (e.g., limiting messages to/from suspects).
• United States v. Zelaya-Veliz: Distinguished because it upheld “two-step” warrants authorizing broad collection but limiting seizure to evidence of enumerated crimes—whereas Bonnell’s was effectively “one-step” (search and seize all data).
• United States v. Charles: Cited via Sardina-Padilla to illustrate federal approval of warrants limited to a “specified, fairly narrow period of time close” to the alleged criminal activity.
• State v. Harris: Reinforced the place-based probable cause concept (fair probability evidence will be found in a particular place), which the court used to critique the warrant’s failure to tailor the search to likely evidence of the homicide.
• State v. Zanter: Used to reject the State’s suggestion that broad warrants should be allowed for authentication/foundation; if a narrower warrant doesn’t yield enough, police should seek a new warrant with additional facts justifying expansion.

These authorities collectively shaped a concrete rule: social media warrants must contain meaningful constraints—temporal and/or subject-matter—when such constraints are feasible given investigative knowledge. A warrant for “all contact info, all photos/videos, all activity logs, all communications, all IP logs” with no time range was treated as the archetypal digital general warrant.

E. Suppression doctrine and harmless-error review

The court’s remedy analysis drew from several lines of authority:

• State v. Malecha and Davis v. United States: Clarified suppression as a prudential doctrine concerning illegally obtained evidence (and corrected the imprecision of “suppressing the warrant”).
• State v. Davis: Supplied the “harmless beyond a reasonable doubt” and “surely unattributable” framework for constitutional errors.
• State v. McDonald-Richards: Provided factors for assessing harmlessness, including cumulativeness, presentation, persuasiveness, use in closing, and whether countered.

On unobjected-to evidence, the court referenced:

• State v. Griller and State v. Patterson: Plain-error framework and substantial-rights requirement.
• Wong Sun v. United States: Fruit-of-the-poisonous-tree concept, though the court declined to reach Bonnell’s alternative taint argument after resolving the warrant issue.

F. Rule 404(b) (immediate episode and Spreigl) and prejudice analysis

The opinion situated the 404(b) debate within Minnesota’s established doctrine:

• State v. Wofford and State v. Riddley: Immediate-episode evidence requires close causal and temporal connection.
• State v. Gomez and State v. Spreigl: Spreigl evidence may be admitted for permitted purposes (motive, intent, identity, etc.).
• State v. Kennedy: Abuse-of-discretion standard for evidentiary rulings.
• State v. Bigbear, State v. Fardan, and State v. Smith: Prejudice framework—whether there is a reasonable possibility the wrongly admitted evidence significantly affected the verdict; strong evidence of guilt diminishes impact.

G. Accomplice corroboration and sufficiency of evidence

The sufficiency discussion tracked long-standing Minnesota rules on accomplice testimony and corroboration:

• State v. Pendelton and State v. Evans: Defined “accomplice” and explained why accomplice testimony is inherently suspect.
• State v. Gilleylen: Articulated the “weighty enough to restore confidence” standard and that corroboration need not cover every element.
• State v. Pederson and State v. Mathiasen: Listed forms corroboration may take (physical evidence; expert testimony; suspicious conduct).
• State v. Thoresen and State v. Rasmussen: Corroboration may be direct or circumstantial.
• State v. Ford and State v. Pippitt: Corroboration evidence is viewed in the light most favorable to the verdict.
• State v. Clark: Distinguished corroboration sufficiency from conviction sufficiency and provided a definition of direct evidence.
• State v. Horst and State v. Webb: Traditional sufficiency review standard when elements are supported by direct evidence.

2) Legal Reasoning

A. The court’s new privacy rule: delivery ends the sender’s privacy interest in the recipient’s copy

The court’s reasoning was functional and control-based. Once an electronic message is received and stored in the recipient’s separate account or device, the sender cannot control whether the recipient saves, forwards, screenshots, publicizes, or voluntarily discloses it. That loss of control renders any continuing privacy expectation objectively unreasonable.

Importantly, the court framed the protected “place searched” as the recipient’s separate account/device and held Bonnell had no legitimate privacy interest in that place. In doing so, the court avoided turning every search of a recipient’s inbox into a derivative search of every sender—a rule that would effectively multiply Fourth Amendment claimants and destabilize settled “delivered communications” doctrine.

B. Standing applied: no ownership interest + no retained privacy expectation = no trigger

Applying the new rule, the court held Bonnell could not invoke the Fourth Amendment or Minnesota Constitution to challenge searches of French’s or Howson’s Facebook accounts. He asserted no ownership interest and no privilege, and the court held he did not retain a reasonable expectation of privacy in recipient-stored copies.

C. A separate and stronger privacy interest exists in a defendant’s own accounts

By contrast, the court recognized that Bonnell did have a reasonable expectation of privacy in data stored in his own Facebook accounts. That distinction cleanly separates (1) the sender’s loss of privacy in the recipient’s copy from (2) the sender’s continuing privacy in his own stored copy and account data.

D. The warrant was invalid because it was functionally unlimited (time and subject)

The warrant sought essentially everything Facebook could provide—contacts, photos/videos, “neoprints,” activity logs, communications, IP logs—without any date range. The court treated the absence of temporal limits as especially decisive in light of State v. Sardina-Padilla, where a three-month limitation made an otherwise broad warrant constitutionally tolerable.

Equally, the court found the warrant lacked meaningful subject-matter tailoring. Investigators knew key facts (they had already interviewed French and Howson and knew communications occurred around January 5–6, 2019), so greater specificity was feasible. The court suggested obvious narrowing options: limiting the timeframe to the weeks leading up to the murder through the arrests; and limiting communications to or from known co-conspirators or otherwise crime-linked content categories.

The court’s critique was not merely formal: in social media accounts, “all content” searches inherently expose vast amounts of unrelated private material. Without constraints, such warrants resemble the general warrants the particularity clause was designed to prevent.

E. Remedy: constitutional violation, but harmless beyond a reasonable doubt

Despite finding a constitutional violation, the court affirmed because the unlawfully obtained evidence from Bonnell’s accounts was largely cumulative of the same messages lawfully obtained from French’s account. The court further concluded that the few exhibits derived exclusively from Bonnell’s accounts did not drive the case: one was minimally used; the others were offered largely to connect Bonnell to the “Pineapple Man” account, a point also supported by other evidence (including photos and phone-number data and additional linking evidence).

The harmlessness conclusion also rested on the substantial independent evidence: French’s eyewitness testimony, incriminating messages in French’s account, DNA evidence linking Bonnell to the gun’s magazine and trigger, and ballistics evidence tying casings in Bonnell’s bedroom to the murder weapon.

F. Rule 404(b): assumed error, no prejudice

The court did not definitively decide whether admitting certain January 2–4 messages under the immediate episode and Spreigl exceptions was correct. Instead, it assumed error and held Bonnell failed to show prejudice: the challenged messages were not central, were limited in use at closing, and were countered by the defense as youthful posturing rather than proof of action.

G. Corroboration and sufficiency: physical evidence and conduct restored confidence in the accomplice eyewitness

The court held corroboration was “weighty enough” because independent evidence linked Bonnell to the gun and the crime and because Bonnell’s messages (lawfully obtained from French’s account) reflected planning and motive and included a post-homicide statement (“I didn’t do it for payment I did it cause I love you”). With corroboration established, French’s eyewitness testimony constituted direct evidence supporting conviction under the traditional sufficiency standard.

3) Impact

A. A bright-line rule for sender privacy in recipient-stored copies

Minnesota now clearly aligns with jurisdictions holding that senders cannot suppress recipient-inbox evidence on the theory that the sender retains privacy in the recipient’s stored copy. Practically, this:

• narrows suppression litigation by limiting who can challenge searches of a given account/device;
• facilitates investigations where accomplices/recipients’ accounts contain key communications;
• shifts defense strategy toward contesting the legality of the account-holder’s search (if the defendant is the account-holder), rather than derivative sender-based claims.

B. Stronger constraints on social-media warrants in Minnesota

The opinion strengthens State v. Sardina-Padilla by clarifying what crosses the line: a warrant for essentially all data in a social media account with no meaningful time range and no subject-matter limitations is unconstitutional. This will likely:

• push investigators to include date ranges tied to known offense windows;
• encourage “two-step” or similarly constrained warrant structures (collection vs. seizure/review constraints) where appropriate;
• increase suppression risk for boilerplate platform warrants that demand all account data “from creation to present.”

C. Harmless error as a limiting principle

Even as the court tightened warrant standards, it signaled that appellate relief will depend on showing the unconstitutional evidence mattered—especially where identical messages exist in constitutionally obtained sources (e.g., recipients’ accounts). In cases with parallel lawful sources, suppression may affect trial strategy and evidentiary presentation but not necessarily the outcome.

Complex Concepts Simplified

Reasonable expectation of privacy

A Fourth Amendment claim generally requires (1) a subjective expectation of privacy and (2) an expectation society recognizes as reasonable. Here, once a message is delivered, the sender’s expectation that the recipient will keep it private is not something the Fourth Amendment protects against government access to the recipient’s stored copy.

“Standing” (who can challenge a search)

Minnesota cases use “standing” as shorthand for whether the defendant had a legitimate expectation of privacy in the place searched. Bonnell could challenge searches of his accounts but not French’s/Howson’s because he had no privacy interest in their accounts and no privacy interest in recipient-stored copies of his messages.

Particularity (why “all content” warrants are dangerous)

The Constitution requires warrants to be specific so police do not rummage through everything. Social media accounts can contain years of unrelated private material. Without time limits or subject limits, a warrant becomes a general warrant—exactly what the Fourth Amendment prohibits.

Temporal limitation and subject-matter limitation

• Temporal limitation: restricting the warrant to a relevant timeframe (e.g., the weeks around the homicide).
• Subject-matter limitation: restricting what categories of data are searched/seized (e.g., messages to/from specific co-conspirators, or content tied to enumerated offenses).

Harmless beyond a reasonable doubt

Even when evidence was admitted in violation of constitutional rights, a conviction can be affirmed if the appellate court concludes the verdict was “surely unattributable” to the error—often because the same evidence was properly admitted elsewhere or because the remaining evidence was overwhelming.

Spreigl and immediate episode (Rule 404(b))

Rule 404(b) generally blocks “prior bad acts” used to suggest a defendant is the kind of person who would commit the charged crime. “Immediate episode” evidence covers acts tightly connected in time/circumstances to tell the full story of the charged offense. “Spreigl” evidence may be admitted for specific purposes (motive, intent, identity), but courts must guard against unfair propensity reasoning.

Accomplice corroboration

Minnesota law does not allow conviction based solely on an accomplice’s testimony. Other evidence must “tend to convict” and be weighty enough to restore confidence in the accomplice’s truthfulness. Corroboration can be circumstantial (DNA, ballistics, suspicious conduct, incriminating statements).

Conclusion

State v. Bonnell establishes two consequential digital-search principles in Minnesota. First, it announces a clear rule that a sender has no reasonable expectation of privacy in the recipient’s stored copy of an electronic message, foreclosing sender-based Fourth Amendment challenges to searches of recipients’ independent accounts. Second, it reinforces and extends Minnesota’s emerging skepticism of broad social media warrants by holding unconstitutional a Facebook warrant for a defendant’s accounts that contained neither temporal nor subject-matter limitations.

At the same time, the decision demonstrates the practical force of harmless-error review in digital evidence cases: even when a defendant wins the constitutional battle on warrant particularity, appellate relief may still turn on whether the same communications were lawfully obtained elsewhere and whether the remaining record independently supports the verdict.