2026 UT App 44

Utah Ct. App.

2026

STATE OF UTAH, Appellee, v. JAMES ERWIN GAINES, Appellant.

No. 20240045-CA

THE UTAH COURT OF APPEALS

Filed March 26, 2026

2026 UT App 44

Second District Court, Ogden Department
The Honorable Cristina P. Ortega
No. 211900210

Emily Adams, Freyja Johnson, and Rachel Phillips Ainscough, Attorneys for Appellant

Derek E. Brown and Marian Decker, Attorneys for Appellee

JUDGE RYAN M. HARRIS authored this Opinion, in which JUDGES GREGORY K. ORME and RYAN D. TENNEY concurred.

HARRIS, Judge:

¶1 Using specialized software known as “Torrential Downpour,” officers got a hit indicating that two specific groups of files of child pornography—or child sexual abuse materials (CSAM)—had been downloaded onto devices linked to James Erwin Gaines. But when officers arrived at Gaines‘s house and searched his devices, they were unable to locate the CSAM files they thought they‘d find. Still, based on the evidence they had from Torrential Downpour, and based on a number of potentially incriminating admissions Gaines made during a police interview, the State charged Gaines with twelve counts of sexual exploitation of a minor. After a trial, a jury convicted Gaines as charged.

¶2 Gaines appeals his convictions, first challenging the trial court‘s denial of his motion to prevent the State from presenting evidence about what Torrential Downpour had indicated related to his devices, as well as his motion for additional discovery related to Torrential Downpour. Gaines also contends that his trial attorney rendered constitutionally ineffective assistance in various respects. While we reject the ineffective assistance claims, we find mеrit in Gaines‘s first argument and conclude that, on this record, the trial court erred in allowing the State to present evidence of Torrential Downpour‘s results.

¶3 But because Gaines admitted that he had knowingly possessed one of the two groups of files, this error harmed Gaines only with respect to the other group of files. We therefore affirm Gaines‘s seven convictions related to the file group he admitted to knowingly possessing, but we reverse his five convictions related to the other file group, and we remand this case to the trial court for further proceedings, including potentially a new trial, on those five counts.

BACKGROUND¹

The Incident

¶4 In April 2019, a police officer (Sergeant) was running the Torrential Downpour² software when he received an alert that it had recorded an IP address that had downloaded some known CSAM, including a group of files collectively named “Tropical Cuties.” Sergeant also learned that the IP address was assigned to a local internet service provider. Sergeant subpoenaed the provider and discovered that the IP address in question was assigned to Gaines. After that, Sergeant did “some investigation” to determine where Gaines lived and “to identify who live[d] in the home” and their “patterns of life,” because if he was “going to serve a warrant,” he wanted Gaines to be home at the time. This investigation continued for about two months.

¶5 During this time, Sergeant continued “running Torrential Downpour” and received notice of “other downloads from that same IP address” that were flagged as CSAM. One of those downloads was a set of files named “Siberian Mouse,” which Sergeant recognized as a “known series” of CSAM. The folder names within the Siberian Mouse files “were all in Russian” and contained photos of “the same female,” who was identified as “Mashka.” Sergeant “viewed each file” and confirmed that the files contained CSAM.

¶6 Sergeant then obtained a search warrant and—accompanied by members of the FBI‘s Child Exploitation Task Force—served that warrant at Gaines‘s house. When Sergeant and the other officers arrived, Gaines answered the door and agreed to speak to Sergeant. After reading Gaines his Miranda rights, Sergeant and another officer (Detective) interviewed Gaines, while the other officers searched the house.

The Police Interview and the Search of Gaines‘s Devices

¶7 During the interview, Detective informed Gaines that he wanted to discuss “the child porn [that was] being downloaded and viewed.” Gaines acknowledged that he had a pornography habit, which he characterized as “disgusting and embarrassing,” but he said it was how he had been “coping with shit.” Gaines acknowledged that he had downloaded quite a bit of pornography, but he claimed it had all been legal post-pubescent pornography, saying that “prepubescent sounds crazy.” He characterized himself as an “ephebophile“—a person who is attracted to post-pubescent but still teenage girls. He clarified, “I will admit that I‘m attracted to girls in the teenage years . . . . [Y]ou know, . . . 17, 16, something like that . . . . But prepubescent, I don‘t have any attraction. That shit‘s just gross. It doesn‘t make sense.”

¶8 Gaines told Sergeant and Detective that during certain times in his life, he had looked at “teen porn,” but he claimed that none of it had been “illegal” and that he was just looking for “the illusion of it, you know, . . . it just says teen.” He stated that he was “absolutely certain” that some of his searches for “teen porn” would “show in [the officers‘] software.” Specifically, Gaines admitted that he had searched for “16Y” images, later explaining that “16Y” meant “[s]ixteen year old or something like that.”

¶9 Gaines also indicated that his knowledge of computers was higher than “the average guy.” He claimed to have “invented software for fighter jets” and to have worked in “IT for ten years before [he] did engineering.” When asked if he used “file sharing” or “peer to peer” software, Gaines responded, “Are you talking about, like, torrents and stuff?” Sergeant answered in the affirmative, and Gaines continued, “Yeah. Torrents, RapidShare, and things like that.”

¶10 Sergeant asked Gaines to explain what he meant when he said he used “peer to peer” and “torrents.” Gaines responded that he understood how torrents worked and, unbidden, said, “Okay. I know what you‘re talking about. This is that . . . Russian thing.” Sergeant asked Gaines to “talk about that,” and Gaines responded, “Okay. Yeah. That one has some bad stuff, and I deleted most of it and kept . . . . [S]o there is a girl, she‘s probably . . . like, 25 now and she still does porn. She did some shit when she was a little younger.” Sergeant asked if Gaines remembered the name of it, and Gaines responded, “Mashka something. And I didn‘t realize, like, how young that went, but I got a porn . . . torrent or whatever with her shit in there, and it had a bunch of stuff. I can‘t read it. It‘s in Russian.” But he acknowledged that, in those files, he “did see some disturbing stuff.” And later in the interview, when discussing “[t]he extent of . . . the child stuff,” Gaines stated, “[T]he only intentional things that I would have on there would be that Russian thing.”

¶11 Sergeant asked if Gaines remembered any files called “Tropical Cuties.” Gaines stated that he knew of a file with that name and had even “searched for it,” but he denied having downloaded it, stating, “I don‘t believe that I did, but it‘s possible I did download it.” He explained, “[S]ometimes you see a cute teenage girl. And, yeah, I do like some of the teenyboppеr, you know, model websites or whatever.” He claimed that what he had downloaded “appeared to be . . . legit,” but that “somehow [he] ended up on one of those torrent websites” and “had a search hit for ‘Tropical Cuties.‘” He claimed that, based on the descriptions of the file, he “didn‘t want to touch it,” and he asserted that he was “still hoping to find” what he called “legit” pornography and that he “didn‘t realize they had other things going on at that company.” In response, Sergeant offered his view that the Tropical Cuties file “ha[d] some pretty clearly prepubescent girls in it,” and Gaines did not disagree, saying, “Yeah, it sounded like it. So . . . unless I was drunk and don‘t remember, I did not download that.” But he later hedged, saying that he “might have” downloaded it but that he had been “hoping” not to.

¶12 Sergeant then asked Gaines to identify the computer he used to “download . . . torrents on,” and Gaines said that the only one would be his laptop computer located in the garage. Sergeant then asked if the hard drive connected to that computer was password-protected, and Gaines confirmed that it was. Gaines told the officers that he didn‘t use the same password for all his devices and that he had “NSA-level encryption on everything.” He indicated that he didn‘t want to share his passwords with the оfficers because in the past “people [had] gotten out of these cases.” Gaines again said, however, that CSAM might be on his devices, stating, “[L]ike I told you, the Mashka one . . . . [S]he‘s attractive. I didn‘t know they went back that young.” He further explained, “I know that I do have at least Mashka, but she‘s not prepubescent. But that torrent came with prepubescent stuff, so that will be there.”

¶13 Eventually, Gaines gave law enforcement some of his passwords and worked with an agent (Agent) to verify that the passwords were accurate. Agent confirmed that he was able to look at Gaines‘s “file history . . . [and] log history” for the laptop in the garage. Agent verified that BitTorrent was on Gaines‘s laptop and noticed the command line “banging teens,” but he did not access that file at that time. Multiple photos were taken of the laptop screen, including a photo of a file with two nude females, one of whom Sergeant and Agent later testified appeared to be “prepubescent.” At that point, while still at Gaines‘s house, Agent “believed . . . he had . . . access” to the laptop and the hard drive, so he “shut it down” and left the scene, intending “to have someone more qualified than [he was] review the data.”

¶14 Later, officers sent the seized devices to a forensics laboratory to be analyzed. At the lab, an analyst discovered that the laptop taken from Gaines‘s garage and its external hard drive could not be examined because they were encrypted and that the passwords Gaines had provided could not bypass the encryption. The analyst was able to view the contents of a small portion of the laptop‘s hard drive but was not able to view the contents of the external hard drive. Sergeant tried to help, and he “attempted multiple passwords without success.” When the analyst tried to bypass the encryption, there was an error message recommending that he back up the data and warning him that, if he proceeded, there was a “high percentage” chance that a “catastrophic loss of data” would occur. The analyst was thus unable to examine the majority of the laptop‘s hard drive or any of the external hard drive. The lab examined the remaining devices and found no additional evidence of CSAM, other than ten “jokes or memes involving [CSAM],” which “were not [CSAM] per se.”

The Motion to Exclude the Torrential Downpour Evidence

¶15 The State eventually charged Gaines with twelve counts of sexual exploitation of a minor—five counts for the Tropical Cuties files and seven counts for the Siberian Mouse files. Each count required the State to prove that Gaines “knowingly . . . possess[ed]” or “intentionally . . . view[ed]” CSAM. Utah Code § 76-5b-201(1) (2019). As the case proceeded toward trial, Gaines‘s attorney (Counsel) filed a motion, grounded in rule 702 of the Utah Rules of Evidence, asking the court to “limit [Sergeant‘s] testimony to the data, principles, and methods that can clearly be shown to the court [to be] reliable.”

¶16 The court held an evidentiary hearing on the motion, where both Sergeant and a supervisory agent (Supervisor) testified about having had previous success using Torrential Downpour despite not knowing how it worked. Sergeant testified to having used Torrential Downpour in “around 30 to 40 [cases] in the past five years,” and he stated that CSAM was usually, but not always, later found in the places where Torrential Downpour indicated it would be found. When asked about how often no CSAM was found, Sergeant responded, “I don‘t know the exact number, but it‘s . . . common enough that it happens.” Sergeant theorized that the reason officers sometimes don‘t find CSAM where expected is that people will “encrypt their devices” or “delete everything.” Supervisor similarly testified to using Torrential Downpour in more than a hundred cases with much success, though there were also times when he didn‘t find CSAM where Torrential Downpour indicated it would be found. Supervisor further stated that he was “sure” that there existed “studiеs to determine [Torrential Downpour‘s] accuracy” but that he was not “aware of any” “[o]ff the top of [his] head.” After the hearing, the court asked both parties to provide additional briefing, and it scheduled oral argument.

¶17 In his brief, Gaines clarified that his challenge was not about what Sergeant “may have seen and recorded” but, instead, went to “the reliability of Torrential Downpour to identify the source of any alleged download as coming from his IP address.” He asserted that neither Sergeant nor Supervisor “could really explain HOW the Torrential Downpour [a]pplication worked,” and he argued that Torrential Downpour is “not reliable enough to determine that any alleged [CSAM] was downloaded or shared by [Gaines] without evidence of that [CSAM] being found on his computer as the result of a search warrant.” The memorandum cited and attached a study on applications like Torrential Downpour, which indicated that those applications had a potential for error. Gaines also submitted an affidavit from a retained expert who was willing to testify that “without any published testing or validation of Torrential Downpour software, [he could not] confirm its reliability, nor [could he] competently prepare [Counsel] to cross-examine any witnesses tеstifying to the use of Torrential Downpour to identify [Gaines‘s] IP address.”

¶18 After full briefing and oral argument, the trial court denied Gaines‘s motion. It acknowledged that Sergeant had testified that he didn‘t “know the software,” but it concluded that Sergeant would be allowed to testify about his use of the software and what the results of the search were. The court analogized Sergeant‘s testimony about Torrential Downpour‘s results to testimony about Intoxilyzer results in DUI cases, and it observed that an “officer . . . trained to operate an Intoxilyzer . . . does not have the knowledge to testify about” the reliability of the machine but could still testify about the results of the test. The court also noted that Gaines‘s motion was limited only to Sergeant‘s testimony, and it thus ruled that the motion was “attacking the wrong person.” The court assured Gaines that he was “more than entitled to bring forth a motion attacking [Torrential Downpour‘s] reliability” but stated that the motion he had filed “misse[d] the mark.” The court ruled that Sergeant was not “excluded as a witness . . . because [he] . . . is trained and can testify as a witness as to his operation of” Torrential Downpour.

The Additional Discovery Motion

¶19 Next, Gaines filed a motion asking the trial court to compel the State to provide him with “a copy of the Torrential Downpour software” as well as “training materials used to train law enforcement in the use of Torrential Downpour,” so that his retained expert could examine its reliability. Gaines asked that the State be required to “disclose . . . any and all evidence favorable to him if the evidence is material to guilt or to punishment,” which he asserted included the requested evidence. (Citing Brady v. Maryland, 373 U.S. 83, 87 (1963).) He cited multiple cases where a criminal defendant was entitled to examine a copy of a government software program “used to investigate child pornography offenses.” He argued that the requested discovery was material “because the [CSAM] files that Torrential Downpour purportedly downloaded from his computer hard drive were never found on any of his electronic devices,” rendering the evidence of what Torrential Downpour had reportedly found on his devices essential to the State‘s case. Gaines thus asserted that “it [was] critical to [his] defense . . . to determine [Torrential Downpour‘s] reliability and accuracy in identifying the files that [he was] charged with.”

¶20 In response, the State argued that the evidence Gaines sought was not material and that—for proprietary reasons—Gaines should not be allowed to discover information regarding Torrential Downpour. In support of its argument, the State cited multiрle cases where federal courts “addressed the discovery of law enforcement investigative software and . . . denied motions similar to [this one].”

¶21 After briefing and argument, the court denied Gaines‘s motion, reasoning that while no CSAM was found on Gaines‘s devices, “a laptop and a hard drive [were] encrypted” and therefore could not “be forensically examined,” and offering its view that “this cuts against [Gaines‘s] argument that the State [was] in possession or could be in possession ‍‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‍of information helpful to his defense.” Counsel had previously argued that “if the whole issue [was] that we need to test” the parts of the laptop and hard drive “that haven‘t been tested yet, . . . we can do that,” but the court still denied the motion, reasoning that Gaines‘s motion was “insufficient to meet the requisite . . . threshold.”

The Trial

¶22 At trial, the State needed to prove beyond a reasonable doubt that Gaines “knowingly . . . possess[ed]” or “intentionally . . . view[ed]” the relevant CSAM files. See Utah Code § 76-5b-201(1) (2019). In support of its case, the State presented testimony from multiple law enforcement officers, admitted multiple photos of Gaines‘s laptop screen, and played nearly an hour of bodycam footage from Sergeant‘s interview with Gaines. During Sergeant‘s testimony, he discussed Torrential Downpour‘s search results, as related to Gaines‘s IP address, and he explained how peer-to-peer sharing worked in this context. During Supervisor‘s testimony, he explained that Torrential Downpour is designed to look for computers that are sharing files known to be CSAM.

¶23 The State also offered testimony, from both Sergeant and Agent, about its Exhibit 35, which was a photo of Gaines‘s laptop screen that depicted two nude females. Both Sergeant and Agent offered their opinions that one of the nude females in the photo appeared to be prepubescent. Gaines was not charged with any crime related to Exhibit 35, but the State argued that because Gaines had Exhibit 35 on his computer, it was more likely that Gaines also had the charged CSAM files on his computer too. Counsel never objected to the officers’ opinions about the age of the female in the photo.

¶24 In his defense, Gaines called his retained expert to the stand, who testified that there were “inconsistencies” in the file sizes from the “log files that [were] generated by Torrential Downpour.” In particular, the expert testified that multiple files in Torrential Downpour‘s log showed that “only a small part” of the file had been “downloaded from the IP address.” This led the expert to hypothesize that Torrential Downpour mаy have erred in linking the CSAM to Gaines‘s IP address.

¶25 In his closing argument, Counsel asserted that Gaines had never actually confessed to downloading CSAM, arguing that Gaines‘s statements about the Siberian Mouse files indicated that he had kept only adult pornography. Counsel also reminded the jury that Gaines had admitted to searching for the Tropical Cuties files but did not admit to downloading those files. In its closing argument, the State reminded the jury that evidence from Torrential Downpour linked the alleged CSAM to Gaines‘s IP address and that Gaines‘s interview with Sergeant contained multiple incriminating admissions, including that he had “intentionally downloaded . . . the Russian file.” It also reminded the jurors that the State had not charged Gaines with any crime related to Exhibit 35.

¶26 After deliberation, the jury convicted Gaines as charged on all twelve counts.

ISSUES AND STANDARDS OF REVIEW

¶27 Gaines now appeals, and he asks us to consider two issues. First, he contends that the trial court erred by allowing the State to present evidence of Torrential Downpour‘s search results. “We review a court‘s decision to admit evidence for abuse of discretion.” State v. King, 2024 UT App 151, ¶ 13, 559 P.3d 96 (cleaned up). “In the event that the trial court admits evidence in error, we will not disturb the outcome of a trial if the admission of the evidence did not reasonably affect the likelihood of a different verdict.” State v. West, 2023 UT App 61, ¶ 15, 532 P.3d 114 (cleaned up). Gaines “bears the burden of showing that [he] was harmed by the trial court‘s error.” Id. (cleaned up).

¶28 Second, Gaines argues that Counsel rendered ineffective assistance. “When an ineffective assistance claim is raised for the first time on appeal, it presents a question of law.” State v. Rivera, 2022 UT App 44, ¶ 21, 509 P.3d 257. Gaines acknowledges that some of his ineffective assistance claims cannot be supported by the evidence in the record, and with regard to these claims, he seeks a remand pursuant to rule 23B of the Utah Rules of Appellate Procedure to enable him to supplement the record with evidence to support the claims. “A remand under rule 23B is available only upon a nonspeculative allegation of facts, not fully appearing in the record on appeal, which, if true, could support a determination that counsel was ineffective.” State v. Tuinman, 2023 UT App 83, ¶ 53, 535 P.3d 362 (cleaned up).

ANALYSIS

I. Torrential Downpour Search Results Evidence

¶29 First, Gaines challenges the trial court‘s decisions denying his two motions regarding the Torrential Downpour search results evidence. Gaines assails the court‘s decision to deny his motion to exclude the evidence as unreliable, as well as the court‘s decision to block him from obtaining further information about Torrential Downpour. We address these challenges in turn.

A. Motion to Exclude Evidence of Search Results

¶30 In addition to contesting Gaines‘s arguments on the merits, the State argues that Gaines failed to preserve for appellate review his challenge to the trial court‘s order denying his motion to exclude the Torrential Downpour evidence. We address the State‘s preservation argument first and reject it, and we then proceed to discuss the merits of Gaines‘s challenge.

1. Preservation

¶31 “Under our preservation rule, any issue brought on appeal must be sufficiently raised to a level of consciousness before the trial court such that the court has an opportunity to rule on it.” State v. Repsher, 2025 UT App 50, ¶ 34, 568 P.3d 1095 (cleaned up). When it comes to preservation of an issue through a motion, “it is the substance, not the labeling, of a motion that is dispositive in determining the character of the motion.” Bishop v. GenTec Inc., 2002 UT 36, ¶ 28, 48 P.3d 218 (cleaned up).

¶32 Here, despite the motion being narrowly titled and at least nominally aimed only at Sergeant‘s testimony, in both the written motion and during oral argument, Gaines sufficiently articulated the true scope of his challenge: that he wanted the trial court to exclude all evidence of Torrential Downpour‘s search results. On the first page of the motion, Gaines requested that “the information allegedly obtained by [officers] resulting from an internet search performed using [Torrential Downpour] be limited to what can be proven reliable.” Gaines repeated this request in his supplemental brief. Later in that same brief, Gaines was clear about the scope of his challenge, stating that he “challenge[d] the reliability of Torrential Downpour to identify the source of any alleged download as coming from his IP address,” and asking that the State “not be allowed to use the evidence of [Sergeant‘s] download at trial.” Then, at the oral argument on the motion, the court made statements indicating that it was aware of the scope of the motion, clarifying with Counsel that Gaines was “concerned about [the software‘s] reliability as far as being able to identify a specific IP address,” and that Gaines was asking for an order forbidding State witnesses from testifying about “the reliability of the identification of the IP address and . . . how it ended up being connected to” Gaines.

¶33 Under these circumstances, Gaines did enough to preserve his challenge. He explained in his written motion and in his supplemental brief what the scope of his request was, thus giving the trial court every opportunity to rule on the challenge.

2. Merits of the Motion

¶34 We therefore proceed to consider the merits of Gaines‘s challenge to the court‘s order denying his motion to exclude the Torrential Downpour search results evidence. Gaines‘s argument for exclusion is grounded in rule 702 of the Utah Rules of Evidence. That rule governs the admissibility of expert testimony, and it requirеs that “a threshold showing” be made “that the principles or methods that are underlying in the testimony (1) are reliable, (2) are based upon sufficient facts or data, and (3) have been reliably applied to the facts.” Utah R. Evid. 702(b). This threshold showing may also be met through an alternative path, “if the underlying principles or methods . . . are generally accepted by the relevant expert community.” Id. R. 702(c).

¶35 With regard to his motion, Gaines asserts that the Torrential Downpour search results evidence was expert testimony within the scope of rule 702; that the State, as the proponent of that evidence, bore the burden of demonstrating that the evidence was reliable; and that the State failed to meet that burden here. In response, the State concedes the first two parts of Gaines‘s argument: that the Torrential Downpour search results evidence falls within the ambit of rule 702 and that the State (as the proponent of the evidence) bears the burden of demonstrating reliability, see State v. Peraza, 2020 UT 48, ¶ 38, 469 P.3d 1023 (“The proponent of the expert testimony has the burden to make [the] threshold showing . . . .” (cleaned up)).

¶36 But the State contests Gaines‘s assertion that it failed to meet that burden. First, the State asserts that—while the trial court did not discuss the rule 702(c) alternative path—we can affirm on that alternative ground here because Torrential Downpour is gеnerally accepted in the relevant expert community. Next, the State asserts that the trial court correctly determined that the rule 702(b) reliability showing was met here. We discuss each of these arguments in turn and conclude that—while Torrential Downpour might well be both reliable and generally accepted—the State failed to actually show either of those things in this case.

a. Rule 702(c)

¶37 The State first asks us to affirm the trial court‘s order on the alternative ground that Torrential Downpour is generally accepted in the relevant community. An appellate court has discretion “to affirm a judgment on an alternative ground if it is apparent in the record.” Olguin v. Anderton, 2019 UT 73, ¶ 20, 456 P.3d 760 (cleaned up). For a ground or theory to be apparent in a record, that record “must contain sufficient and uncontroverted evidence supporting the ground or theory to place a person of ordinary intelligence on notice that the prevailing party may rely thereon on appeal.” Id. (cleaned up). In this case, the State‘s proffered alternative ground is not apparent in the record.

¶38 During the briefing and argument process associated with Gaines‘s motion to exclude, the State never argued—or tried to prove—that Torrential Downpour was generally accepted in the relevant expert community. By contrast, Gaines argued in his trial court briefs that Torrential Downpour was not generally accepted in the relevant community, and—in resisting Gaines‘s motion to exclude—the State did not attempt to counter this argument or present evidence to the contrary. Nevertheless, here on appeal the State points to testimony from Sergeant and Supervisor indicating that they used Torrential Downpour regularly and were trained on its use and operation, and the State argues that “[l]aw enforcement‘s repeated use of Torrential Downpour in [its] peer-to-peer cases fits squarely within rule 702(c) because it shows [that] the software is generally accepted by the relevant ‍‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‍expert community—peer-to-peer investigators.” This is simply not enough for us to conclude that Torrential Downpour‘s general acceptance is apparent on the record in this case.

¶39 For purposes of our discussion, we will assume—without deciding—that the relevant expert community for evaluating the general acceptance of Torrential Downpour consists of law enforcement officers who investigate CSAM crimes. But even indulging this assumption, the record submitted to us does not sufficiently support the cоnclusion that this community, as a whole, generally accepts Torrential Downpour as reliable. In this case, no witness was asked whether Torrential Downpour is generally accepted in any community of law enforcement officers, and no witness offered any testimony on the subject. While two officers from Utah testified that they regularly use the software, no testimony was given about, for instance, whether those officers are required to use Torrential Downpour (as opposed to other software applications), how often other officers use Torrential Downpour, or whether such use is limited to these officers’ department or to the State of Utah (as opposed to used nationally).

¶40 If Torrential Downpour really is generally accepted in the law enforcement community, it shouldn‘t be difficult to make that showing. The problem here is that no such showing was made, and we are not comfortable concluding by implication, merely from testimony about two Utah officers’ regular use, that Torrential Downpour is generally accepted in the community of CSAM-enforcement officers. We therefore decline the State‘s invitation to affirm, on this alternative ground, the trial court‘s conclusion that Torrential Downpour is reliable.

b. Rule 702(b)

¶41 Next, thе State asserts that it made the necessary threshold showing by meeting the criteria of rule 702(b). Specifically, it defends the trial court‘s determination that evidence about Torrential Downpour‘s search results was reliable and based upon sufficient facts or data that have been reliably applied to the facts. See Utah R. Evid. 702(b). The trial court‘s conclusion was based largely on testimony from Sergeant and Supervisor that they regularly use Torrential Downpour and that they usually—but not always—find CSAM where Torrential Downpour indicates that it will be found. From this testimony, the trial court concluded that the threshold reliability showing had been made and that the State would be able to present evidence—largely through Sergeant—that Torrential Downpour had located CSAM at an IP address registered to Gaines.

¶42 Gaines assails this conclusion, and he takes the position that, in order to make the threshold reliability showing, the State needed to present testimony about the inner workings of Torrential Downpour from a software expert who could explain how the program works. As Gaines sees it, because the State did not present any such testimony, it failed to make the required threshold showing of reliability.

¶43 We see shortcomings in the positions of both parties. We disagree with Gaines‘s assertion that the only way to show that Torrential Downpour is reliable is to present testimony from a software expert. But we also disagree with the State‘s assertion that it did enough here to make the threshold showing.

¶44 It seems to us that the State had several options at its disposal through which it could have made the required threshold showing that Torrential Downpour reliably identifies CSAM at specific IP addresses.³ First, it could have done as Gaines

believes is required by presenting testimony from a software expert who could have explained how Torrential Downpour works and who could have potentially testified, based on the software science underlying the program, that its results are reliable. The State took a similar approach in State v. Roberts, 2015 UT 24, 345 P.3d 1226. In that case, law enforcement officers used a software program known as “the Wyoming Toolkit” to monitor peer-to-peer file-sharing networks for CSAM. Id. ¶ 3. At an evidentiary hearing, the State called “an expert on the Wyoming Toolkit” to testify about its “methodologies“; this expert also opined “that the Toolkit correctly identifies [CSAM] files with extraordinarily high accuracy.” Id. ¶ 13. Based on this evidence, the district court determined that the Wyoming Toolkit was reliable, and the Utah Supreme Court affirmed that ruling. Id. ¶¶ 55–56. In this case, the State did not avail itself of that option; it did not attempt to demonstrate Torrential Downpour‘s reliability through the testimony of a software expert. But this was not the only option open to the State.

¶45 Alternatively, the State could have attempted to present evidence of scientific studies demonstrating Torrential Downpour‘s reliability and accuracy. Indeed, Counsel asked Supervisor about this during the evidentiary hearing. Specifically, Counsel asked whether Supervisor knew about “any kind of studies done to determine [Torrential Downpour‘s] accuracy,” and Supervisor responded by stating that he was “sure” that such studies must have been done but that he was not aware of any

results are generally not needed or admitted to prove, at trial, that the defendant had CSAM on his or her devices. Because the CSAM files that Torrential Downpour identified were ultimately not found on Gaines‘s devices, this case presents the unusual situation in which the Torrential Downpour search results form a critical and substantive part of the State‘s evidence at trial. Thus, this case is one of those presumably rare cases in which it ends up being necessary for the State to demonstrate the reliability of Torrential Downpour‘s search results.

“[o]ff the top of [his] head.” But the State did not follow up on this line of questioning, and it made no effort to introduce evidence of any specific studies that attempted to quantify the accuracy of Torrential Downpour‘s search results.

¶46 Despite presenting no testimony from a software expert and no evidence of any pertinent studies, the State nevertheless argues that it made the necessary threshold reliability showing by presenting testimony of Sergeant and Supervisor, who each testified to regular use of Torrential Downpour, and who each testified that, in their experience, Torrential Downpour is accurate most of the time. But these two officers’ use of Torrential Downpour does not represent a large enough sample size to draw meaningful conclusions about the software‘s reliability over time; indeed, the officers were not able to offer specific percentages of success, and they acknowledged that, at least some of the time, CSAM files ended up not being located where Torrential Downpour indicated they would be found. Sergeant even stated that such instances were “common enough.” To be sure, the officers explained away these instances, asserting that sometimes the devices are encrypted and can‘t be searched, and that sometimes the suspect is presumed to have deleted the files. But this fact—that, in somе percentage of cases significant enough to be referred to as “common,” the CSAM files identified by Torrential Downpour are not found on the devices associated with the suspected IP address—is a problem for the State in this context. And in our view, anecdotal testimony from two officers that they regularly used Torrential Downpour and that it seemed to work most of the time is just not enough to meet the requirements of rule 702(b).

¶47 To be clear, we are not willing to categorically hold that a proponent of Torrential Downpour search results evidence could never make the required rule 702(b) reliability showing through “use” evidence. Perhaps such a showing would be possible with a better evidentiary record than this one. (Although such evidence might be better suited toward demonstrating general acceptance in the relevant community, pursuant to rule 702(c).) We hold merely that the State failed to meet the required showing based on the evidence it presented in this case.

¶48 We are also unpersuaded by the State‘s attempt to analogize this case to Eskelson ex rel. Eskelson v. Davis Hospital & Medical Center, 2010 UT 59, 242 P.3d 762. In that case, a plaintiff sued a doctor who allegedly perforated his child‘s eardrum while trying to extract an object lodged in the child‘s ear. See id. ¶ 1. The plaintiff retained an expert witness whose opinion—that the doctor had breaсhed the standard of care—“relied heavily” on the child‘s mother‘s deposition and did not “identify a scientific methodology beyond his expertise as a physician.” Id. ¶¶ 13–14. The district court excluded the testimony as unreliable under rule 702, but our supreme court reversed, emphasizing that the requirements of the threshold reliability showing “will vary depending on the complexity of the particular case,” and holding that the plaintiff‘s expert had sufficient “experience with the removal of foreign objects from the ears of children” to satisfy “the threshold showing that his testimony was reliable.” Id. ¶ 15. In the State‘s view, Eskelson is “determinative” here and compels the conclusion that the State has met the threshold showing through the officers’ use testimony.

¶49 We disagree. At issue there was the relatively simple procedure of removing an object from an ear. In that situation, an expert with experience with that procedure was allowed to testify, even without referencing a specific methodology, that the defendant doctor had breached the standard of care. Id. What‘s at issue here, by contrast, is much more complex: the reliability of the results of a seemingly complicated software program. And the witnesses the State chose to testify about the results do not understand how the software actually works аnd have no information about its success rate over time and over adequate sample sizes. We agree with Gaines that Eskelson is materially distinguishable from the facts of this case.

¶50 We also find unhelpful the trial court‘s effort to analogize Sergeant‘s and Supervisor‘s testimony here to officers’ testimony, in other cases, about Intoxilyzer results. To be sure, most officers who testify about Intoxilyzer results likely don‘t know the inner workings of how the machine functions. But Intoxilyzer machines have been around a long time and have long been considered “universal[ly] accept[ed]” as reliable. See Murray City v. Hall, 663 P.2d 1314, 1320 (Utah 1983). Moreover, our legislature has passed a statute creating “a presumption that the test results are valid and further foundation for introduction of the evidence is unnecessary.” Utah Code § 41-6a-515(3). On appeal, the State directs our attention to no such statute regarding Torrential Downpour, nor does it direct our attention to any judicial opinions discussing the reliability of Torrential Downpour. And in the absence of better evidence regarding Torrential Downpour‘s reliability, we agree with Gaines that the more apt analogies here are to things like polygraph technology or artificial intelligence—applications that often yield useful results but which are not (or at least not yet) considered dependable enough to have thеir results be routinely admitted as evidence against a criminal defendant.

¶51 In the end, we agree with Gaines that, on this record, the State did not present evidence sufficient to meet its burden of making the threshold rule 702(b) showing of reliability regarding the Torrential Downpour search results. Accordingly, we find merit in Gaines‘s challenge to the trial court‘s decision to allow testimony about those search results. On this record, that testimony should not have been admitted.

c. Prejudice

¶52 Because we have concluded that evidence of the Torrential Downpour search results was erroneously admitted at trial, we must now address the ‍‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‍question of prejudice: whether Gaines was harmed by this error. “Not every trial error regarding the admission of evidence requires reversal.” In re A.D.-C., 2024 UT App 150, ¶ 33, 559 P.3d 984. “Even if the evidence was improperly admitted, we will not overturn a conviction unless a reasonable likelihood exists that the error affected the outcome of the proceedings.” State v. Bilek, 2018 UT App 208, ¶ 19, 437 P.3d 544 (cleaned up). In assessing harm, we engage in a “counterfactual” inquiry, and we “consider a hypothetical—an alternative universe in which the trial went off without the error“—and ask whether a different outcome would have been reasonably likely in that counterfactual alternative trial. See State v. Ellis, 2018 UT 2, ¶ 42, 417 P.3d 86. “A reasonable likelihood requires a рrobability sufficient to undermine confidence in the outcome.” State v. Klenz, 2018 UT App 201, ¶ 64, 437 P.3d 504 (cleaned up).

¶53 As we have already mentioned, the Torrential Downpour search results evidence was an important part of the State‘s case against Gaines. After all, officers did not find the CSAM files on Gaines‘s devices that Torrential Downpour indicated they‘d find. So the State‘s case on the twelve CSAM counts depended largely on the Torrential Downpour evidence, as well as on the potentially incriminating admissions Gaines made during his interview with officers. As we discuss, the admissions Gaines made regarding the Siberian Mouse files were more significant than those he made regarding the Tropical Cuties files; indeed, the Siberian Mouse admissions were all but determinative. Accordingly, we conclude that the trial court‘s error was harmless as it concerns the Siberian Mouse files, but not as it concerns the Tropical Cuties files.

¶54 We begin our prejudice analysis by refocusing on the terms of the relevant statute. To obtain a conviction on the CSAM counts, the State had to prove that Gaines either (a) “knowingly . . . possess[ed]” or (b) “intentionally . . . view[ed]” the CSAM files referenced in each count. See Utah Code § 76-5b-201(1) (2019).

¶55 With regard to the Siberian Mouse files, Gaines‘s own admissions make plain that he knowingly possessed CSAM. Gaines brought up what he called the “Russian thing” unbidden, before even being asked about it. He admitted that some of the files Sergeant was looking for were named “Mashka something” and that he knew they contained “some bad stuff.” He explained that, regarding “child stuff,” “the only intentional things that [he] would have on there would be that Russian thing.” And when telling Sergeant what he expected law enforcement to find on his laptop, he acknowledged that he had “at least Mashka” on there and that he was aware that “that torrent came with prepubescent stuff, so that will be there.” While Gaines did not admit to actually viewing the “prepubescent” portions of those files, he admitted that those files were on his laptop and that he knew they contained CSAM. Through these statements, Gaines effectively admitted to knowingly possessing CSAM in the form of the Siberian Mouse files. So when we envision a counterfactual trial in which the jury heard these admissions but did not hear about the Torrential Downpour search results, we discern no reasonable likelihood of a different outcome with regard to the charges concerning the Siberian Mouse files.

¶56 Our conclusion is different, however, with regard to the Tropical Cuties files. As to those files, Gaines‘s admissions—while certainly not ideal from his perspective—were not nearly as damning. In his interview, Gaines admitted to searching for Tropical Cuties and to getting “a search hit” for Tropical Cuties, but he claimed that, after he read the torrent‘s description, he decided not to “touch it” because it may have had CSAM and he was looking for what he called “legit” porn. Sergeant then told Gaines that the Tropical Cuties file had “some pretty clearly prepubescent girls in it,” to which Gaines responded, “Yeah, it sounded like it. So . . . unless I was drunk and don‘t remember, I did not download that.” Thus, Gaines did not admit to either knowingly possessing or intentionally viewing the Tropical Cuties files. So with regard to these files, the State‘s primary evidence was the Torrential Downpour search results indicating that the files would be found on devices linked to Gaines‘s IP address. In a counterfactual trial in which the Torrential Downpour evidence was excluded, there is a reasonable likelihood of a different result. Accordingly, the trial court‘s error in admitting the Torrential Downpour evidence was not harmless as to those files.

¶57 In sum, then, the court‘s error prejudiced Gaines with regard to the five counts related to the Tropical Cuties files. On this basis alone, we reverse Gaines‘s convictions on those five counts, and we remand this matter to the trial court for further proceedings on those counts.

¶58 But because the court‘s error in admitting the Torrential Downpour evidence was harmless with regard to the seven counts related to the Siberian Mouse files, we must proceed to consider—or at least discuss—Gaines‘s remaining arguments.

B. Motion for Additional Discovery

¶59 Gaines‘s other challenge related to the Torrential Downpour search results evidence concerns the trial court‘s denial of his motion asking the court to compel the State to produce additional information about Torrential Downpour. In that motion, Gaines asked for “a copy of the Torrential Downpour software” as well as “training materials used to train law enforcement in the use of Torrential Downpour,” so that his retained expert could examine its reliability. Given our conclusions regarding Gaines‘s challenge to the court‘s admission of the Torrential Downpour search results evidence, see supra Part I.A.2, we need not make an ultimate ruling on the merits of Gaines‘s related challenge to the court‘s denial of his discovery motion.⁴ But because these issues may resurface on remand

related to a retrial regarding the Tropical Cuties counts, we nevertheless elect to discuss some of the issues Gaines raises, in an effort to provide guidance to the trial court. See State v. Low, 2008 UT 58, ¶ 61, 192 P.3d 867 (stating that, when “there are other issues presented on appeal that will likely arise” on remand, we may “exercise our discretion to address those issues for purposes of providing guidance on remand“).

¶60 Specifically, we elect to address Gaines‘s contention that the trial court erred in concluding that the additional information Gaines sought regarding Torrential Downpour was not material to Gaines‘s defense. In the court‘s words, it was “not convinced how the requested discovery [would] significantly help” Gaines‘s defense. This ruling was incorrect, given the way the issues were presented to the trial court.

¶61 During criminal prosecutions, the State is obligated to disclose “all evidence favorable to the defendant that is material to guilt or punishment.” Utah R. Crim. P. 16(a)(1)(F). “Evidence is material if there is a reasonable probability that, had the evidence been disclosed to the defense, the result of the proceeding would have been different.” Tillman v. State, 2005 UT 56, ¶ 29, 128 P.3d 1123 (cleaned up).

¶62 As already noted, see supra note 3, in many CSAM cases information about Torrential Downpour could turn out not to be material, because in most such cases CSAM is actually found on the defendant‘s devices. In those cases, the evidence that serves as the centerpiece of the prosecution‘s case is the fact that CSAM was located on the defendant‘s devices; evidence of what led officers, in the first place, to suspect that CSAM would be found on those devices is of diminished relevance.

¶63 But in cases like this one, where CSAM files were not located on the defendant‘s devices, and where the State‘s case depends, in significant part, on evidence of Torrential Downpour‘s search results, additional information about Torrential Downpour may well turn out to be material. In this context, the materiality inquiry will depend on several variables, including (for instance) what specific additional information is sought and which method (e.g., software expert, studies, widespread usage) the prosecuting entity intends to use in its attempt to meet rule 702‘s threshold reliability requirements. We of course do not know exactly how this issue will present itself on remand, should the State elect to retry Gaines on the Tropical Cuties counts and should Gaines elect to renew his discovery motion. We are thus unable to offer any opinion on whether any such renewed motion should (or should not) be granted.

¶64 However, in this case, denial of such a motion cannot be supported solely by the argument that Torrential Downpour evidence is immaterial because officers—due to encryption and lack of passwords—were unable to exhaustively search Gaines‘s devices for CSAM. It is, of course, the State‘s burden to prove Gaines guilty beyond any reasonable doubt. Here, that requires the State to show that Gaines either “knowingly . . . possess[ed]” or “intentionally . . . view[ed]” CSAM. See Utah Code § 76-5b-201(1) (2019). And given Gaines‘s constitutional right against self-incrimination, “the government must ‘produce [its] evidence . . . by its own independent labors’ and may not extract such evidence from a person ‘by the cruel, simple expedient of compelling it from his own mouth.‘” See State v. Fullerton, 2018 UT 49, ¶ 17, 428 P.3d 1052 (quoting Miranda v. Arizona, 384 U.S. 436, 460 (1966)). The trial court seemed to hold it against Gaines, when assessing the materiality of the additional Torrential Downpour information Gaines was seeking, that Gaines had not made more of an effort to facilitate the State‘s search of his encrypted devices. Yet the State made no argument below, and makes no argument now, that Gaines was obligated to make any such effort.

¶65 Thus, should Gaines renew his discovery motion on remand, the court should consider it with this guidance in mind and should not deny it merely because Gaines did not voluntarily provide additional passwords to his encrypted devices.

II. Ineffective Assistance of Counsel

¶66 Next, Gaines argues that Counsel rendered ineffective assistance. To succeed on such a claim, Gaines must make a two-part showing. He “must show that [C]ounsel‘s performance was deficient,” which “requires showing that [C]ounsel made errors so serious that [C]ounsel was not functioning as the ‘cоunsel’ guaranteed the defendant by the Sixth Amendment.” Strickland v. Washington, 466 U.S. 668, 687 (1984). And he “must show that the deficient performance prejudiced the defense,” which “requires showing that [C]ounsel‘s errors were so serious as to deprive [him] of a fair trial, a trial whose result is reliable.” Id. Gaines must prove both elements. Id. “Because failure to establish either prong of the test is fatal to an ineffective assistance of counsel claim, we are free to address [such claims] under either prong.” State v. Popp, 2019 UT App 173, ¶ 25, 453 P.3d 657 (cleaned up). And if a claim is infirm under one of the prongs, then “the claim fails and the court need not address the other.” State v. Nelson, 2015 UT 62, ¶ 12, 355 P.3d 1031.

¶67 To establish deficient performance, Gaines must show that Counsel‘s “representation fell below an objective standard of reasonableness.” Popp, 2019 UT App 173, ¶ 26 (cleaned up). In evaluating the reasonableness of an attorney‘s actions, courts will often look to whether the actions the attorney took were motivated by trial strategy. See State v. Scott, 2020 UT 13, ¶ 35, 462 P.3d 350 (“To be sure, the performance inquiry will often include an analysis of whether there could have been a sound strategic reason for [the attorney‘s] actions.“). And while “the ultimate question is not whether there was a possible strategic reason for [the attorney‘s] conduct, but instead whether that conduct was objectively reasonable,” id., “[i]f it appears [that the attorney‘s] actions could have been intended tо further a reasonable strategy, a defendant has necessarily failed to show unreasonable performance,” State v. Ray, 2020 UT 12, ¶ 34, 469 P.3d 871.

¶68 The second part of the test requires Gaines to show that he was prejudiced by Counsel‘s performance. “Prejudice exists when there is a reasonable probability that the case would have had a different outcome had trial counsel not performed deficiently.” State v. Whytock, 2020 UT App 107, ¶ 28, 469 P.3d 1150. “A reasonable probability is a probability sufficient to ‍‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‍undermine confidence in the outcome” of the proceeding. Strickland, 466 U.S. at 694. In assessing prejudice, we “consider the totality of the evidence before the judge or jury and then ask if the defendant has met the burden of showing that the decision reached would reasonably likely have been different absent the errors.” State v. Garcia, 2017 UT 53, ¶ 28, 424 P.3d 171 (cleaned up).

¶69 As already explained, see supra ¶ 52, “[p]rejudice analysis is counterfactual.” State v. Ellis, 2018 UT 2, ¶ 42, 417 P.3d 86. “To decide whether a trial affected by error is reasonably likely to have turned out differently we have to consider a hypothetical—an alternative universe in which the trial went off without the error.” Id.; see also State v. Soto, 2022 UT App 107, ¶ 25, 518 P.3d 157 (“Under a counterfactual analysis, we consider whether, in the absence of the improperly admitted evidence, the likelihood of a different outcome is sufficiently high to undermine our confidence in the verdict.” (cleaned up)).

¶70 In this case, Gaines asserts that Counsel rendered ineffeсtive assistance in four different ways: by (1) not making a motion for a directed verdict based on insufficient evidence; (2) not impeaching Agent‘s testimony at trial with his report; (3) not playing more of the video of Gaines‘s police interview for the jury; and (4) not objecting to Sergeant‘s and Agent‘s opinions about the age of a female in a photo (Exhibit 35) taken of Gaines‘s laptop screen. Gaines asserts that claims (1) and (4) can be supported with evidence already in the record, but he acknowledges that evidence supporting claims (2) and (3) is not already in the record. On those claims, Gaines has filed a motion, pursuant to rule 23B of the Utah Rules of Appellate Procedure, asking us to remand this case to the trial court for further proceedings aimed at supplementing the record with evidence to support these claims. For the reasons discussed below, however, we reject each of Gaines‘s ineffective assistance arguments and deny his rule 23B motion because Gaines has not demonstrated that Counsel rendered constitutionally ineffective assistance.⁵

A. Directed Verdict

¶71 First, Gaines argues that Counsel rendered ineffective assistance by not making a motion for a directed verdict asserting that the State had failed to present sufficient evidence to support a conviction on the Siberian Mouse counts. We reject this argument because Gaines has not demonstrated that Counsel performed deficiently by opting not to make such a motion.

¶72 We begin our analysis by reiterating three relevant principles. First, “a directed verdict is warranted only when, viewed in the light most favorable to the State, no evidence exists from which a reasonable jury could find beyond a reasonable doubt that the defendant committed the crime.” State v. Pola, 2025 UT App 143, ¶ 35, 579 P.3d 407 (cleaned up), cert. denied, 581 P.3d 560 (Utah 2025). Second, “the decision not to pursue a futile motion is almost always a sound trial strategy.” State v. Orton, 2024 UT App 140, ¶ 25, 558 P.3d 443 (cleaned up). And third, “the propriety of directed verdict rulings must be evaluated on the basis of the evidence actually presented, including even evidence that is later, on appeal, determined to have been inappropriately admitted.” State v. Hoffman, 2021 UT App 143, ¶ 21 n.6, 503 P.3d 538; cf. Franklin v. Stevenson, 1999 UT 61, ¶ 7, 987 P.2d 22 (“Whether competent or incompetent, all evidence submitted to the jury must be considered by the court in ruling on a motion for judgment notwithstanding the verdict.” (cleaned up)). Thus, even though we‘ve concluded that the Torrential Downpour evidence was improperly admitted, we must consider that evidence as part of the universe of material that the trial court would have been allowed to consider when ruling on any directеd verdict motion.

¶73 So here, not only could the jury consider Gaines‘s damning admissions about the Siberian Mouse files, see supra ¶ 55, it could also consider the Torrential Downpour search results evidence. With all of this evidence in the record, there was more than enough evidence to support a conviction on the Siberian Mouse counts. And under these circumstances, a reasonable attorney could have determined not to make a directed verdict motion on those counts. Accordingly, Gaines has not demonstrated that Counsel performed deficiently by opting not to do so.

B. Impeachment

¶74 Second, Gaines argues that Counsel was ineffective for opting not to impeach Agent‘s testimony with Agent‘s written investigative report. At trial, Agent suggested that Exhibit 35—a photo taken of Gaines‘s laptop screen, discussed more fully below—showed a nude female who appeared to be prepubescent, but nowhere in Agent‘s report did he note that he had found CSAM on Gaines‘s computer, instead concluding that “further review was needed.” Agent‘s report was never introduced at trial, so Gaines asks us to remand the case to the trial court to facilitate supplementation of the record with the report.

¶75 Under the Utah Rules of Appellate Procedure, a defendаnt “may move the court to remand the case to the trial court for entry of findings of fact, necessary for the appellate court‘s determination of a claim of ineffective assistance of counsel.” Utah R. App. P. 23B(a). However, “if the defendant could not meet the test for ineffective assistance of counsel, even if his [or her] new factual allegations were true, there is no reason to remand the case, and we should deny the motion.” State v. Griffin, 2015 UT 18, ¶ 20, 441 P.3d 1166.

¶76 Here, Counsel did not render ineffective assistance by opting not to use Agent‘s report to impeach Agent‘s trial testimony on this point. Competent counsel could have reasonably concluded that attempting to impeach Agent in this manner would have been unsuccessful. Agent‘s report concluded that “further review was needed,” and if Counsel had tried to impeach Agent with this report, Agent very likely would have clarified that, after further review, he had determined that Exhibit 35 depicted an underage nude female, a sequence of events that could have served to further emphasize Agent‘s opinion.

¶77 But even if Counsel‘s decision in this regard could somehow be considered deficient performance, Gaines has not shown that this performance prejudiced him on the Siberian Mouse counts. As discussed, the evidence against Gaines on those counts was overwhelming given his admissions. See supra ¶ 55. Thus, Gaines‘s ineffective assistance claim in this regard fails on both parts of the Strickland test.

C. Contextualizing Gaines‘s State of Mind

¶78 Third, Gaines asserts that Counsel was ineffective for electing not to admit additional portions of Gaines‘s interview that he now claims could have “contextualized” his “state of mind” and helped the jury to understand that he was not “a pedophile trying to cover his tracks” but was, instead, “someone [just] struggling to cope with a multitude of difficulties in his life.” In particular, he faults Counsel for not telling the jury, among other things, that Gaines‘s wife had been diagnosed with schizoaffective disorder and that Gaines had been having serious financial struggles. He claims that evidence of these stressors would have served to put his other more incriminating statements into context and to help explain his state of mind.

¶79 Counsel did not perform deficiently by electing not to inform the jury about these additional contextual facts. Counsel could reasonably have recognized that any benefit to Gaines from presenting this evidence would be minimal at best. The reasons Gaines might have chosen to download and view CSAM are potentially relevant to sentencing but have little relevance during the guilt phase. A jury convinced that Gaines downloaded CSAM would hardly be any less likеly to convict him simply by learning that he was experiencing various life struggles.

¶80 And again, even if Counsel could somehow be deemed to have performed deficiently in this regard, Gaines has not shown that this performance prejudiced him on the Siberian Mouse counts. As discussed, the evidence against Gaines on those counts was overwhelming. See supra ¶ 55. Thus, Gaines‘s third ineffective assistance claim fails on both parts of the test.

D. Exhibit 35

¶81 Finally, Gaines argues that Counsel was ineffective for not objecting to Sergeant‘s and Agent‘s testimonies about the age of one of the nude females depicted in Exhibit 35. Notably, Gaines does not argue that Exhibit 35 was inadmissible; he limits his admissibility challenge to the officers’ opinions about the age of the female in the photo. In particular, he points out that, at least in cases “where it is difficult to tell whether the individual depicted is older or younger than eighteen,” expert testimony about the individual‘s age is “necessary to help the trier of fact reach a reasoned conclusion.” See State v. Jordan, 2018 UT App 187, ¶ 59, 438 P.3d 862 (cleaned up). The State does not take issue with this general proposition, but it asserts that Sergeant and Agent—as law enforcement officers with significant experience in CSAM investigations—are the sort of experts who are qualified to offer opinions in this contеxt. Gaines resists this conclusion, and he takes the position that individuals with medical training are the only ones qualified to offer an admissible expert opinion about the age of an individual in an alleged CSAM image. The parties’ dispute on this point presents an interesting legal issue that Utah appellate courts, to our knowledge, have yet to confront.

¶82 But we need not address that legal issue here, because even if we assume, without deciding, that Counsel performed deficiently by not objecting to Sergeant‘s and Agent‘s opinions, Gaines has failed to show that this deficient performance prejudiced him. Here, as already noted, our prejudice analysis is confined only to the Siberian Mouse counts, and on those counts the evidence against Gaines—even without the Torrential Downpour search results evidence and even without Exhibit 35—was overwhelming. As already noted, see supra ¶ 55, Gaines admitted to possessing the Siberian Mouse files and to knowing that they contained CSAM. On this basis, we reject Gaines‘s ineffective assistance claim related to the officers’ opinions regarding the age of the individual depicted in Exhibit 35.

CONCLUSION

¶83 We find merit in Gaines‘s challenge to the trial court‘s decision denying his motion to exclude evidence of the Torrential Downpour search results, because the State failed to make the threshold showing that such evidence was reliable. And this error was not harmless regarding the five counts related to the Tropical Cuties files. Accordingly, we reverse Gaines‘s convictions on those five counts, and we remand this case to the trial court for further proceedings on those counts.

¶84 But the trial court‘s error was harmless as to the seven counts related to the Siberian Mouse files. And we reject all of Gaines‘s other appellate arguments, including his ineffective assistance claims and the arguments made in his 23B motion (which we deny). Accordingly, we affirm Gaines‘s convictions on the seven Siberian Mouse counts.

Notes

“In an appeal from a jury trial, we review the record facts in a light most favorable to the jury‘s verdict and recite the facts accordingly, and we present conflicting evidence only as necessary to understand issues raised on appeal.” (cleaned up).

This case involves several terms with which readers may not be familiar, including “BitTorrent,” “torrent,” and “Torrential Downpour.” As explained by witnesses at trial, “BitTorrent” is a peer-to-peer file-sharing program available to anyone on the internet. A “torrent” is a file or folder that is downloaded through use of a file-sharing network, such as BitTorrent. And “Torrential Downpour” is а law-enforcement-specific version of BitTorrent that allows law enforcement to become one of the “peers” and search for known CSAM files on peer devices.

Although some law enforcement officers investigating CSAM crimes apparently regularly use Torrential Downpour, our understanding is that the State is not often in the position of having to demonstrate the reliability of this software. We suspect that this is because, in most similar cases the State files, CSAM ends up actually being found on the target devices during a search. In such cases, Torrential Downpour‘s search results are used, at most, to show probable cause for a search warrant; those (continued…)

Because we have reversed Gaines‘s convictions on the Tropical Cuties counts, Gaines‘s challenge to ‍‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‌‍the court‘s discovery order is moot as to those counts. And for the reasons already explained, see supra Part I.A.2.c, due to his incriminating admissions, Gaines would have almost certainly been convicted of the Siberian Mouse counts in a counterfactual trial in which the jury heard none of the Torrential Downpour search results evidence; thus, any error the court might have made in denying Gaines‘s discovery motion is also harmless as to the Siberian Mouse counts.

Given our ruling, see supra Part I.A, reversing Gaines‘s convictions on the Tropical Cuties counts, our analysis here concerns only the Siberian Mouse counts.

Read the detailed case summary