Arapahoe County District Court Case No. 24CR53 Honorable David N. Karpel, Judge

          Attorneys for Plaintiff-Appellant:

          Amy L. Padden, District Attorney, Eighteenth Judicial District

          Laura Wood, Deputy District Attorney

          Centennial, Colorado

          Attorneys for Defendant-Appellee:

          Tucker Grosgebauer, P.C.

          Grant W. Grosgebauer

          Jason C. Fisher

          Denver, Colorado

          JUSTICE HOOD delivered the Opinion of the Court, in which CHIEF JUSTICE MARQUEZ, JUSTICE BOATRIGHT, JUSTICE GABRIEL, JUSTICE HART, JUSTICE SAMOUR, and JUSTICE BERKENKOTTER joined.

          OPINION

          HOOD JUSTICE

         ¶1 Defendant, Patrick Nkongolo, has been charged with multiple counts of sexual assault on a child, A.K., as a pattern of abuse. In this interlocutory appeal, the prosecution challenges the trial court's pretrial order, which suppressed a textmessage conversation between Nkongolo and A.K.'s father, D.K., that occurred on November 15, 2023. The trial court concluded that the statements Nkongolo made during that conversation were the product of police coercion, which rendered them involuntary and inadmissible at trial.

         ¶2 D.K. initiated the conversation at the behest of law enforcement, and a police officer guided D.K. through the questioning. So, it's undisputed that D.K. acted as an agent of the police during the November 15 conversation. We conclude, however, that D.K.'s conduct wasn't coercive. And even if it had been, this alleged coercion didn't play a significant role in inducing Nkongolo's side of the text exchange. Consequently, we reverse the portion of the trial court's order that suppressed Nkongolo's November 15 statements.

         I. Facts and Procedural History

         ¶3 In 2023, A.K. told a therapist that Nkongolo had repeatedly sexually assaulted her over the previous three years, beginning when she was eleven years old. A.K. referred to Nkongolo as her uncle, even though they are not related, and explained that he was a close family friend who had lived with her family for

several years. The therapist reported the outcry to Arapahoe County Human Services ("ACHS"), and a case worker at ACHS reported it to the police.

         ¶4 As part of the police investigation into the allegations, the investigating officer asked D.K. to initiate a "pretext conversation" with Nkongolo by text message. As the officer later explained at the pretrial hearing, a pretext conversation is "an investigative tool, to see how a suspect is going to respond to involved parties in the case." The officer testified that she told D.K. ahead of time what type of messages she wanted him to send, and during the conversation, she made suggestions to D.K. about what to say and, more specifically, what questions to ask.

         ¶5 D.K.'s first attempt at the pretext conversation on November 2 was unsuccessful. Nkongolo sent D.K. a message a few days later, on November 7, but D.K. said he couldn't talk. Finally, on November 15, the pretext conversation at issue occurred.^[1]

         ¶6 At the beginning of the conversation, D.K. told Nkongolo that he wanted to talk about what had happened with A.K. He then told Nkongolo, "[Y]ou are a

member of the family, you must tell me sincerely that it is happening to see what we can do." D.K. let Nkongolo know that A.K. had already spoken to him, but said he wanted to hear from Nkongolo. Nkongolo downplayed what had occurred, explaining it involved nothing more than hugs and jokes. But D.K. said that didn't align with what A.K. had told him and that he "need[ed] the truth to see what we can do to keep this in the family." When Nkongolo again said it was just a hug, D.K. confronted him with A.K.'s accusation that Nkongolo had put his "mouth [o]n her breasts." Nkongolo's version remained the same. Although, during this conversation, Nkongolo never explicitly admitted to any unlawful sexual contact, he eventually apologized for upsetting A.K. and admitted to giving her "a little friendly kiss" when she hugged him.

         ¶7 D.K. asked Nkongolo several more times whether he had kissed A.K.'s breasts, imploring Nkongolo to be honest, but Nkongolo never confirmed or denied that allegation. Finally, D.K. ended the conversation by saying, "I wanted this to be dealt with in the family but apparently you don't want to so I'm going to do what [my wife] wants us to do."

         ¶8 The prosecution subsequently charged Nkongolo with five counts of sexual assault on a child as a pattern of abuse. Before trial, Nkongolo moved to suppress all three of the November text exchanges.

         ¶9 At the hearing on the motion, the trial court heard testimony from the investigating officer and reviewed the text messages. It then concluded that the statements Nkongolo made on the first two dates, November 2 and 7, were voluntary and admissible, and it denied suppression of those statements.

         ¶10 The trial court also concluded, however, that the prosecution had failed to show by a preponderance of the evidence that the November 15 statements were voluntary. The court found that D.K.'s repeated assertions that he wanted to "keep this in the family" were implied promises that D.K. wouldn't go to the police if Nkongolo confessed to what he'd done. The court found that these repeated implied promises were coercive and that the totality of the circumstances overbore Nkongolo's will, rendering his statements involuntary. So, the court granted Nkongolo's motion to suppress those statements.

         ¶11 The prosecution now appeals that ruling.

         II. Jurisdiction and Standards of Review

         ¶12 We may consider an interlocutory appeal filed by the prosecution, seeking relief from a trial court's suppression order. § 16-12-102(2), C.R.S. (2024); C.A.R. 4.1(a); People v. Brown, 2022 CO 11, ¶ 13, 504 P.3d 970, 974.

         ¶13 Because a suppression ruling presents a mixed question of law and fact, we "defer to the trial court's findings of historical fact, if supported by competent evidence in the record," but "we review the trial court's conclusions of law de

novo." Brown, ¶ 14, 504 P.3d at 975. So here, we defer to the trial court's factual findings regarding the circumstances surrounding Nkongolo and D.K.'s conversation, but we consider anew whether those circumstances rendered Nkongolo's statements involuntary. See People v. Matheny, 46 P.3d 453, 461-62 (Colo. 2002); People v. McIntyre, 789 P.2d 1108, 1111 (Colo. 1990).

         ¶14 "If the defendant makes a prima facie showing of involuntariness at a suppression hearing, the prosecution then bears the burden of establishing by a preponderance of the evidence that he in fact made the statements voluntarily." People in Int. of Z.T.T., 2017 CO 48, ¶ 11, 394 P.3d 700, 703 (quoting People v. McIntyre, 2014 CO 39, ¶ 15, 325 P.3d 583, 587). And although we must consider the totality of the circumstances, People v. Davis, 187 P.3d 562, 563-64 (Colo. 2008), on appeal "we look solely to the record created at the suppression hearing" to determine whether the trial court properly suppressed the evidence, People v. Thompson, 2021 CO 15, ¶ 16, 500 P.3d 1075, 1078.

         III. Analysis

         ¶15 "Under the due process clauses of the United States and Colorado constitutions, a defendant's statements must be voluntary to be admissible as evidence." People v. Ramadon, 2013 CO 68, ¶ 18, 314 P.3d 836, 841; see also U.S. Const. amends. V, XIV; Colo. Const. art. II, § 25. These constitutional requirements exist even if the defendant was not in custody when the statements were made.

People v. Medina, 25 P.3d 1216, 1225 (Colo. 2001); see also Colorado v. Connelly, 479 U.S. 157, 163-67 (1986). This is to say nothing more remarkable than determining whether the safeguards established in Miranda v. Arizona, 384 U.S. 436 (1966), apply and were followed and determining whether a statement was voluntary are different analyses, even though the relevant legal considerations are intertwined. As a result, even though a statement by phone doesn't constitute custodial interrogation, People v. Platt, 81 P.3d 1060, 1066 (2004), the trial court was still obligated to address the voluntariness of Nkongolo's statements, see id. ("Statements and confessions received as a result of a non-custodial interrogation are admissible, if they are voluntary.").

         ¶16 "To be voluntary, a statement must be the product of an essentially free and unconstrained choice by its maker." Ramadon, ¶ 19, 314 P.3d at 842. A defendant's statements are therefore involuntary if an officer's coercive conduct played a significant role in overbearing the defendant's will and inducing the statements. Z.T.T., ¶ 12, 394 P.3d at 703.

         ¶17 Thus, an involuntary statement has three attributes. First, we must identify some form of governmental coercion. Blum v. Yaretsky, 457 U.S. 991, 1004 (1982); see also Connelly, 479 U.S. at 166 ("The most outrageous behavior by a private party seeking to secure evidence against a defendant does not make that evidence inadmissible under the Due Process Clause."). State action may seem obvious

when the allegedly coercive actors are uniformed police officers but less so when law enforcement uses a private party to interrogate a suspect.

         ¶18 Second, we consider whether the state actor's conduct was actually coercive; meaning, sufficiently forceful to implicate constitutional limits on governmental interrogation. Connelly, 479 U.S. at 167. Coercive conduct may include physical abuse or threats as well as more "subtle forms of psychological coercion." Effland v. People, 240 P.3d 868, 877 (Colo. 2010). But context is king. So, "we examine 'both the defendant's ability to resist coercive pressures and the nature of the police conduct.'" Z.T.T., ¶ 13, 394 P.3d at 703 (quoting Ramadon, ¶ 20, 314 P.3d at 842). The following non-exhaustive list of factors helps guide this determination:

1. whether the defendant was in custody;

2. whether the defendant was free to leave;

3. whether the defendant was aware of the situation;

4. whether the police read Miranda rights to the defendant;

5. whether the defendant understood and waived Miranda rights;

6. whether the defendant had an opportunity to confer with counsel or anyone else prior to or during the interrogation;

7. whether the statement was made during the interrogation or volunteered later;

8. whether the police threatened [the] defendant or promised anything directly or impliedly;

9. the method [or style] of the interrogation;

10. the defendant's mental and physical condition just prior to the interrogation;

11. the length of the interrogation;

12. the location of the interrogation; and

13. the physical conditions of the location where the interrogation occurred.

Id. (alterations in original) (quoting McIntyre, ¶ 17, 325 P.3d at 587).

         ¶19 Third, even if the conduct was coercive, a final step remains. Id. at ¶ 12, 394 P.3d at 703. For us to conclude that a defendant's statement was involuntary, we must determine that coercive police conduct played a "significant role" in inducing it. Id.

         ¶20 With these legal concepts in mind, we return to the case at hand.

         A. Government Actor

         ¶21 Because D.K. was not a police officer, we must first determine whether he was acting as an agent of the government in eliciting Nkongolo's statements. We begin by considering "(1) whether the government 'encourage[d], initiate[d], or instigate[d] the private action,' and (2) whether 'the party performing the [interrogation] intended to assist law enforcement efforts or to further his own ends.'" People v. Pilkington, 156 P.3d 477, 479 (Colo. 2007) (first three alterations in original) (quoting United States v. Smythe, 84 F.3d 1240, 1242-43 (10th Cir. 1996)).

         ¶22 Here, D.K. contacted Nkongolo at law enforcement's behest, and an officer told D.K. which questions to ask. The trial court found that D.K. was an agent of the police, and neither party disputes this finding.

         B. Coercive Conduct

         ¶23 Next, we must consider whether D.K.'s conduct was coercive. We examine coercion from the suspect's perspective. Illinois v. Perkins, 496 U.S. 292, 296 (1990).

         ¶24 We recognize that the "police-dominated atmosphere" of a prototypical custodial interrogation generates "inherently compelling pressures which work to undermine the individual's will to resist and to compel him to speak where he would not otherwise do so freely." Id. (quoting Miranda, 384 U.S. at 445, 467); accord Matheny, 46 P.3d at 462-63. But these "inherently compelling pressures" are missing when an individual speaks to a friend; that is, when an individual doesn't know he is speaking with the police. Perkins, 496 U.S. at 296-97 (quoting Miranda, 384 U.S. at 467). Therefore, it isn't inherently coercive for police officers to use an agent, like D.K., to try to get a suspect to speak. See id. at 297 (explaining that "mere strategic deception by taking advantage of a suspect's misplaced trust" in someone the suspect doesn't know is an agent of the police doesn't offend due process). Accordingly, we must consider the totality of the circumstances and the remaining factors to determine whether Nkongolo's statements were voluntary.

         ¶25 First, Nkongolo wasn't in custody during the November 15 conversation. Although D.K.'s questions constituted an interrogation, see Rhode Island v. Innis, 446 U.S. 291, 301 (1980), Nkongolo wasn't subjected to a custodial interrogation. Thus, he wasn't entitled to a Miranda warning. See People v. Wakefield, 2018 COA 37, ¶ 47, 428 P.3d 639, 649. These circumstances therefore weigh against a finding of coercion. See Effland, 240 P.3d at 874.

         ¶26 Second, Nkongolo arguably had the opportunity to speak with an attorney before this conversation but chose not to. He told D.K. that he had received a letter from ACHS with a number he could call, but he wanted to speak with D.K. first. As the trial court observed, however, it's unclear that Nkongolo knew he could speak with an attorney or even that he might want to. So, this factor doesn't tilt the scales one way or the other.

         ¶27 Third, we note that the nature of the conversation-text messaging-weighs strongly against coercion. See People v. Munoz-Diaz, 2023 COA 105, ¶ 15, 543 P.3d 402, 406. Nkongolo could choose the timing, his location, and the duration of the conversation. He could also choose to respond or to stop responding at any time. And nothing about these messages indicates that Nkongolo was mentally or physically impaired.

         ¶28 Lastly, D.K. told Nkongolo at the beginning of the conversation that he wanted to talk about what had happened with A.K., so Nkongolo was aware of

the situation. D.K. also said A.K. had told him what happened, but he wanted to hear it from Nkongolo so they could "see what we can do." D.K. repeatedly said that he wanted Nkongolo to be truthful and that he wanted "to keep this in the family." After Nkongolo failed to answer D.K.'s questions about whether he had kissed A.K.'s breasts, D.K. ended the conversation by saying, "I wanted this to be dealt with in the family but apparently you don't want to so I'm going to do what [my wife] wants us to do."

         ¶29 The trial court found that D.K.'s statements about wanting "to keep this in the family" were implied promises that if Nkongolo told D.K. the truth, D.K. wouldn't go to the police. Because the record supports this finding, we defer to it. We disagree, however, with the trial court's legal conclusion that these circumstances were coercive.

         ¶30 We generally don't consider it coercive for law enforcement officers to tell a suspect the factual allegations against him. People v. Cerda, 2024 CO 49, ¶ ¶ 42-43, 559 P.3d 206, 215. Nor do we generally consider it coercive for law enforcement officers to tell a suspect the possible consequences that may result from his decision to speak or to remain silent. People v. Smiley, 2023 CO 36, ¶ 39, 530 P.3d 639, 649; see also 2 Wayne R. LaFave et al., Criminal Procedure § 6.2(c) n.100, Westlaw (4th ed. database updated Nov. 2024) ("[A] mere threat to take action which would be lawful and necessary absent cooperation is not objectionable."); United States v. Perez,

127 F.4th 146, 173 (10th Cir. 2025) (explaining that courts generally don't consider it coercive to accurately explain the reality of a suspect's situation and the consequences of cooperation versus silence). And "[t]he practice of encouraging a suspect to be honest is well-established . . . as noncoercive conduct." United States v. Pena, 115 F.4th 1254, 1263 (10th Cir. 2024); see also People v. Miranda-Olivas, 41 P.3d 658, 662-63 (Colo. 2001).

         ¶31 Therefore, considering the totality of the circumstances surrounding the November 15 conversation, we conclude that the prosecution failed to prove that D.K.'s conduct was coercive. See Z.T.T., ¶¶ 14-15, 394 P.3d at 703-04; Munoz-Diaz, ¶¶ 15-21, 543 P.3d at 406-07.

         C. Significant Factor

         ¶32 Finally, even if D.K.'s conduct had been coercive, we would still need to determine whether this conduct played a significant role in inducing Nkongolo's statements. The trial court concluded that D.K.'s implied promises not to go to the police, combined with "the fact that it's unknown whether . . . Nkongolo[] was aware of the situation" because the contents of the letter from ACHS weren't in the record, were circumstances that played a significant role in overbearing Nkongolo's will. Accordingly, the court concluded Nkongolo's statements were involuntary. We disagree.

         ¶33 An officer's repeated exhortations for a suspect to be honest don't necessarily overbear a defendant's free will. See Pena, 115 F.4th at 1264. Similarly, an officer may generally "make a truthful statement regarding a possible punishment without it overbearing a defendant's will," Dowell v. Lincoln Cnty., 762 F.3d 770, 776 (8th Cir. 2014), as long as the officer isn't intentionally exploiting a suspect's known weaknesses with such statements, see Ramadon, ¶¶ 24-28, 314 P.3d at 844-45.

         ¶34 Here, there is no evidence that Nkongolo was particularly vulnerable to D.K.'s implied promises that he would keep the situation in the family and not involve the police if Nkongolo told him the truth or that these promises were an attempt to exploit some weakness. And Nkongolo never changed his story, maintaining throughout the conversation that he had just been joking around. So, it's clear that neither D.K.'s implied promises, nor the circumstances as a whole, overbore Nkongolo's will or induced him to speak. Cf. People v. Springsted, 2016 COA 188, ¶¶ 34-49, 410 P.3d 702, 712-15 (concluding that the coercive environment created by the officers played a significant role in overbearing the defendant's will, as evidenced, in part, by the fact that the defendant changed his story as the coercive tactics mounted).

         ¶35 We therefore conclude that Nkongolo's statements were voluntary.

         IV. Conclusion

         ¶36 We reverse the portion of the trial court's order suppressing the November 15 statements, and we remand the case to the trial court for further proceedings.

---------

Notes:

^[1] Nkongolo and D.K. are originally from the Democratic Republic of the Congo and primarily speak Tshiluba. In text messages, however, they typically use French. The investigating officer had the transcripts of the three relevant text exchanges translated from French to English, which the court admitted into evidence at the suppression hearing. We rely on that translation in this opinion.

---------