Peo v. Gibbens-Martinez

Opinion

23CA1390 Peo v Gibbens-Martinez 09-03-2026 COLORADO COURT OF APPEALS

Court of Appeals No. 23CA1390 Adams County District Court No. 22CR2110 Honorable Jane A. Tidball, Judge

The People of the State of Colorado, Plaintiff-Appellee, v. Ashley Shelee Gibbens-Martinez, Defendant-Appellant.

JUDGMENT AFFIRMED

Division IV

Opinion by JUDGE WELLING

Bernard*, J., concurs

Lum, J., concurs in part and dissents in part

NOT PUBLISHED PURSUANT TO C.A.R. 35(e)

Announced September 3, 2026

Philip J. Weiser, Attorney General, Grant R. Fevurly, Senior Assistant Attorney General, Denver, Colorado, for Plaintiff-Appellee

Megan A. Ring, Colorado State Public Defender, Andrew Vierra, Deputy State Public Defender, Denver, Colorado, for Defendant-Appellant

*Sitting by assignment of the Chief Justice under provisions of Colo. Const. art. VI, § 5(3), and § 24-51-1105, C.R.S. 2026.

¶1 Defendant, Ashley Shelee Gibbens-Martinez, appeals her judgment of conviction for second degree aggravated motor vehicle theft. We affirm.

I. Background

¶2 While at a motel, Officer Marc Faivre observed a white Kia Sorento in the parking lot. He ran the license plate and saw that it had been reported stolen. He approached the car and saw a woman, later identified as Gibbens-Martinez, in the driver’s seat. Officer Faivre observed that because the driver’s side door handle was broken, he was unable to open the door from the outside. He also observed that the ignition had been damaged so that it was possible to start the car without a key.

¶3 Officer Faivre spoke with Gibbens-Martinez. She told him that earlier that day she had borrowed the car from a friend whom she had known for just a few weeks. When he asked her about the damage to the door handle, she said that she hadn’t noticed it. She, however, had noticed the damage to the ignition, which had led her to ask her friend if the car had been stolen. Gibbens- Martinez told Officer Faivre that she had used a nail file to start the

2

car, which she said she had previously used to start her own car once when her boyfriend had taken her keys.

¶4 Gibbens-Martinez was arrested and charged with one count of second degree aggravated motor vehicle theft, which includes as an element that the car was worth $20,000 or more at the time it was stolen. At trial, her theory of defense was that she didn’t know that the car had been stolen. The People didn’t argue that Gibbens- Martinez had originally stolen the car. Instead, they argued that because of the circumstances under which she came into possession of the car and the damage to it, she knew that she was exercising control over a stolen car. The jury convicted her as charged, and the trial court sentenced her to three years of community corrections.

II. Issues on Appeal

¶5 Gibbens-Martinez raises three arguments on appeal. She contends that (1) the trial court erred by admitting a Kelley Blue Book (KBB) valuation report of the car because (a) the prosecution failed to establish an adequate foundation for its admission and (b) its admission violated the Confrontation Clause; (2) the prosecution committed misconduct during voir dire and closing

3

argument; and (3) the trial court erred by using the 2022 model criminal jury instruction’s definition of reasonable doubt. She also contends that the cumulative effect of these errors requires reversal. We consider, and reject, each contention in turn below.

A. Kelley Blue Book Valuation Report Admissibility

¶6 Gibbens-Martinez contends that the trial court reversibly erred by admitting a KBB report of the car’s value. We disagree.

1. Additional Facts

¶7 During Officer Faivre’s testimony at trial, the People sought to admit a report from KBB, a website used to estimate the value of cars based on a variety of factors, including the make, model, year, and mileage of the car. The proffered report, which is set forth below, purported to contain an estimated valuation of a 2018 Kia Sorento with 40,000 miles.

4

(KBB Report for 2018 Kia Sorento)

¶8 Before the prosecution moved to admit the report, Officer Faivre testified that the year, make, and model of the car described in the report matched that of the car he found in Gibbens- Martinez’s possession.

¶9 Defense counsel made a hearsay objection, and during the ensuing bench conference, the court also questioned whether the

5

People had laid an adequate foundation for the document to be admitted. The People explained that while Officer Faivre wasn’t the officer who had generated the KBB report, he had laid foundation for its admissibility by “testifying that the make, model, and year as listed on that Kelley Blue Book valuation is the same make, model, and year of the vehicle he located the defendant in.” The People admitted that they didn’t know who had generated the report, just that it was an officer other than Officer Faivre. Defense counsel responded that “[f]or a Kelley Blue Book valuation, there’s a lot of other selections that are involved. Someone would have had to input that.” The trial court overruled defense counsel’s objection, explaining that any concern that Officer Faivre wasn’t the officer who had generated the report went to weight, not admissibility. At no point during this discussion did anyone mention the Confrontation Clause.

2. Authenticity

¶ 10 First, Gibbens-Martinez argues that Officer Faivre didn’t lay an adequate foundation to establish the report’s authenticity. We aren’t persuaded that the trial court abused its discretion by finding that he had.

6

a. Standard of Review and Applicable Law

¶ 11 We review a trial court’s evidentiary ruling for an abuse of discretion. Gonzales v. People, 2020 CO 71, ¶ 25. “A court abuses its discretion when its ruling is manifestly arbitrary, unreasonable, or unfair, or it is based on a misunderstanding or misapplication of the law.” People v. Claycomb, 2025 COA 36, ¶ 14. “The requirement of authentication or identification as a condition precedent to admissibility is satisfied by evidence sufficient to support a finding that the matter in question is what its proponent claims.” CRE 901(a). CRE 901 doesn’t establish a specific set of criteria for authenticating evidence; rather, the CRE 901 standard is a flexible one. People v. N.T.B., 2019 COA 150, ¶¶ 17, 33.

¶ 12 Thus, the burden a proponent must meet to authenticate a document “‘is not high — only a prima facie showing is required,’ and a ‘district court’s role is to serve as gatekeeper in assessing whether the proponent has offered a satisfactory foundation from which the jury could reasonably find that the evidence is authentic.’” People v. Glover, 2015 COA 16, ¶ 13 (citation omitted). Once the proponent meets this burden, the actual authenticity of

7

the evidence and the effect of any defects go to the weight of the evidence and not its authenticity. N.T.B., ¶ 16.

b. Analysis

¶ 13 Gibbens-Martinez argues that Officer Faivre (1) couldn’t authenticate the KBB report because he didn’t generate it and had no personal knowledge about who did and (2) failed to authenticate the report because he didn’t testify about how the report was generated, including what data was entered into the KBB website to generate the report. We disagree and conclude that the trial court didn’t abuse its discretion by finding that these concerns went to the weight the jury should accord the KBB report, not the report’s authenticity.

¶ 14 As noted, to satisfy CRE 901, the People needed to show that the document is what it’s purported to be — that is, it’s a KBB report estimating the value of a particular make, model, and year of a car. Officer Faivre testified that the year, make, and model of the car in the report matched the car he found Gibbens-Martinez in. This testimony was sufficient, albeit minimally, to authenticate the report (and provide support for the report’s relevance). Accordingly, the trial court didn’t abuse its discretion in concluding that the

8

report was what the prosecution claimed it to be — a KBB report for a 2018 Kia Sorento. See People v. Abad, 2021 COA 6, ¶ 45 (concluding that the trial court didn’t abuse its discretion in finding that an officer who didn’t generate cell phone reports still authenticated the reports through his testimony by showing that the reports were what the prosecution claimed they were (data downloaded from a cell phone extraction)); see also Gonzales, ¶ 42 (“[T]he standard for authentication is minimal — all that’s required is a prima facie showing that the evidence is what its proponent claims.”).

¶ 15 As for Gibbens-Martinez’s concerns about how the report was created, including what data was entered to generate the report, those concerns go to weight and potentially relevancy, not authenticity. For example, if Gibbens-Martinez was concerned that the condition of the car was worse than what was used to generate the report — which would have meant that the value of the stolen car was lower than what was reflected in the report — she could have asked what condition of the car the report was based on or if the report accounted for the condition Officer Faivre found the car in, including the damage to the driver’s side door and ignition. She

9

didn’t cross-examine Officer Faivre about this or any of the concerns she raises about the reliability of the report to establish the value of the stolen car. All this, as well as Officer Faivre’s familiarity (or lack of familiarity) with the report, go to the weight the fact finder should have given the report, not to its authenticity or admissibility. See People v. Thornton, 251 P.3d 1147, 1150 (Colo. App. 2010) (rejecting the defendant’s argument that “because the officer did not testify sufficiently to the data he entered into the Blue Book’s website, the resulting valuation was incompetent and inadmissible” because “this argument goes only to the weight and not the admissibility of the valuation”); cf. Robinson v. People, 927 P.2d 381, 384 (Colo. 1996) (explaining that the witness’s degree of familiarity with the defendant goes to the weight to be given to the witness’s identification of the defendant, not the admissibility of such testimony).

¶ 16 Accordingly, we conclude that the trial court didn’t abuse its discretion by finding that the People had adequately authenticated the KBB report through Officer Faivre’s testimony.

10

3. Confrontation Clause

¶ 17 Second, Gibbens-Martinez argues that the admission of the KBB report violated her rights under the Confrontation Clauses of the United States Constitution and the Colorado Constitution. U.S. Const. amends. VI, XIV; Colo. Const. art. II, § 16. We perceive no reversible error.

a. Standard of Review and Applicable Law

¶ 18 The United States and Colorado Constitutions guarantee criminal defendants the right to confront the witnesses against them. See U.S. Const. amend. VI; Colo. Const. art. II, § 16. But these protections apply only to testimonial statements. Nicholls v. People, 2017 CO 71, ¶ 33. A testimonial statement is one made “under circumstances that would lead an objective witness reasonably to believe that the statement would be available for use at a later trial.” Id. at ¶ 22 (citing Crawford v. Washington, 541 U.S. 36, 51-53 (2004)).

¶ 19 Gibbens-Martinez doesn’t dispute, and we agree, that she didn’t preserve her Confrontation Clause objection to the KBB report, so our standard of reversal is plain error. See Hagos v. People, 2012 CO 63, ¶ 14; see also People v. Crabtree, 2024 CO

11

40M, ¶ 27 (“[F]or errors that have not been preserved by objection (whether constitutional or unconstitutional), the case-specific, outcome-determinative standard of reversal we apply is plain error review.”). To show plain error, a “defendant must establish that [an] error occurred, that the error was obvious, and that the error’s effect is so grave that it undermines the fundamental fairness of the trial itself and casts doubt upon the reliability of the conviction.” People v. Ambrose, 2021 COA 62, ¶ 66 (citation omitted).

b. Analysis

¶ 20 Gibbens-Martinez argues that the KBB report violated her Confrontation Clause rights because it was a testimonial report and the officer who generated the report didn’t testify. We disagree and conclude that the report wasn’t testimonial, and, in any event, any asserted error wasn’t obvious.

¶ 21 To begin, we reject any argument that the KBB website’s underlying data is testimonial. That data wasn’t compiled to aid in a criminal prosecution, and it isn’t meaningfully different than a printed copy of the KBB. See Nicholls, ¶¶ 22, 33; see also People v. Hard, 2014 COA 132, ¶ 15 n.3 (concluding that the information on Drugs.com, a website used to identify prescription drugs, “is not

12

testimonial for purposes of the Sixth Amendment’s Confrontation Clause” because it wasn’t created primarily for the purpose of establishing facts relevant to later criminal prosecution); Walker v. Commonwealth, 704 S.E.2d 124, 126 (Va. 2011) (explaining that a National Automobile Dealers Association “blue book” wasn’t testimonial because it wasn’t compiled for the purpose of securing the defendant’s conviction).

¶ 22 Whether a KBB report generated after an officer enters data about a specific car into the KBB website is testimonial may be a closer call. But tellingly, Gibbens-Martinez doesn’t cite any authority, and we haven’t found any, supporting the proposition that a KBB report generated for litigation is testimonial. Indeed, the weight of authority persuades us that the report isn’t testimonial when, like here, the underlying data isn’t testimonial. See, e.g., People v. Ortega, 2016 COA 148, ¶ 15 (concluding that a record custodian’s distillation of phone records into an exhibit offered at trial didn’t transform the records into testimonial statements); People v. Marciano, 2014 COA 92M-2, ¶ 40 (concluding that duplicates of bank statements obtained in the course of investigating the case weren’t testimonial).

13

¶ 23 In any event, given the lack of authority supporting Gibbens- Martinez’s contention, any error isn’t obvious and, therefore, isn’t plain. See People v. Vigil, 251 P.3d 442, 447 (Colo. App. 2010) (holding that an alleged error isn’t plain or obvious given the lack of authority supporting the defendant’s contention).

B. Prosecutorial Misconduct

¶ 24 Gibbens-Martinez next contends that reversal is required because of prosecutorial misconduct during voir dire and closing argument. We aren’t persuaded.

1. Legal Principles and Standard of Review

¶ 25 When reviewing a claim of prosecutorial misconduct, we engage in a two-step analysis. Wend v. People, 235 P.3d 1089, 1096 (Colo. 2010). First, we determine whether the prosecutor’s challenged conduct was improper based on the totality of the circumstances and, second, whether such conduct warrants reversal applying the appropriate standard of reversal. Id.

¶ 26 Gibbens-Martinez didn’t object to any of the alleged misconduct, so we review for plain error. See id. at 1097; Hagos,

¶ 14. Prosecutorial misconduct constitutes plain error if it’s “‘flagrant or glaringly or tremendously improper’ and so

14

undermine[s] the fundamental fairness of the trial as to cast serious doubt on the reliability of the judgment of conviction.” People v. Carian, 2017 COA 106, ¶ 52 (quoting People v. Cevallos-Acosta, 140 P.3d 116, 122 (Colo. App. 2005)).

2. Voir Dire

¶ 27 Gibbens-Martinez argues that the prosecutor committed misconduct by using voir dire to educate the jury on the prosecution’s theory of the case and to pre-try the case. We perceive no plain error.

a. Additional Facts

¶ 28 As noted above, the prosecution’s theory of the case was that Gibbens-Martinez knew that she was exercising control over a stolen car because of the circumstances under which she came into possession of the vehicle and the car’s damage that was readily apparent. During his voir dire, the prosecutor posed the following questions about a hypothetical borrowed lawn mower to prospective jurors:

• Would you feel comfortable borrowing a lawn mower from a neighbor?

15

• What factors would you consider when deciding whether you felt comfortable borrowing a lawn mower from your neighbor?

• Would you feel comfortable borrowing a lawn mower that had a security tag on it from your neighbor?

• Would you feel comfortable borrowing a lawn mower from someone you hadn’t met before?

• Would you feel comfortable borrowing a lawn mower from someone who happened to be walking by when you were asking your neighbor to borrow a lawn mower?

¶ 29 During this questioning, prospective jurors mentioned several concerns that they had with borrowing a lawn mower under the hypothesized circumstances, such as how well they knew the person they were borrowing the lawn mower from or whether they thought the lawn mower was stolen. The prosecutor responded that those concerns were “valid” and “interesting.” The prosecutor summed up these concerns with the following statement to the venire:

[T]hroughout this conversation, we’ve really been talking about kind of two factors or dimensions, but the first is the thing itself and

16

whether it looks like it’s in good condition and whether it looks like it’s on the up-and-up versus whether it has a security tag or whether there is some indication that this thing is not on the up-and-up and might not actually belong to the person who’s lending it. That’s the first thing we’ve been talking about. And the second thing we’ve been discussing is who is the person lending it to you and how well do you know that person.

¶ 30 The prosecutor then asked the potential jurors if anyone “would have some concern about borrowing that lawn mower with the security tag on it?” In response to this question, a potential juror said they would be comfortable borrowing it but “[i]f it was something different, you know, a car, that’s different.” The prosecutor then asked the jury venire, “[Who is] cool borrowing a vehicle from someone they didn’t know . . . ?” He followed up by asking what are some reasons that they wouldn’t feel comfortable borrowing a car from someone they didn’t know.

¶ 31 Towards the end of his voir dire, the prosecutor presented another hypothetical, asking whether any of the prospective jurors would buy a lawn mower that had a security tag on it from someone selling lawn mowers outside a home improvement store. When a juror responded that they wouldn’t buy the lawn mower because

17

they would assume that person had “received the product without paying for it,” the prosecutor responded, “Fair. Thank you.” For another potential juror, the prosecutor asked if part of their concern was that the lawn mower could be stolen. The prosecutor ended his conversation with that juror by asking if they would agree with him that if they were concerned that the person was selling stolen goods, then the person wasn’t authorized to sell those goods. The prospective juror answered, “Fair enough.”

b. Analysis

¶ 32 Gibbens-Martinez argues that the prosecutor used these questions and hypotheticals to improperly educate the jury on the prosecutor’s theory of the case by telling the potential jurors what factors were important to deciding the case. We perceive no plain error.

¶ 33 “A prosecutor engages in prosecutorial misconduct during voir dire when she misstates the law or ‘intentionally use[s] the voir dire to present factual matter which the prosecutor knows will not be admissible at trial or to argue the prosecution’s case to the jury.’” People v. Krueger, 2012 COA 80, ¶ 50 (citation omitted); see People v. Carter, 2015 COA 24M-2, ¶ 71. “In determining whether

18

prosecutorial misconduct has occurred, ‘[t]he context in which [the] challenged prosecutorial remarks are made is significant.’” Krueger,

¶ 50 (alterations in original) (citation omitted).

¶ 34 Gibbens-Martinez doesn’t cite, and we haven’t found any, cases where a reviewing court in Colorado has reversed when a prosecutor has improperly pre-tried his case or argued his case to the jury, much less reversed under a plain error standard of reversal. Indeed, in Krueger, a division of this court analyzed a prosecutor’s comments during voir dire about a nontestifying participant that couldn’t be forced to testify and concluded that the comments were proper because “the prosecutor never explicitly referred to [the nontestifying participant’s] privilege against self- incrimination.” Id. at ¶ 55. Based on this, we can’t conclude that any asserted error was obvious or flagrant, especially given that the prosecutor didn’t tell the prospective jurors how the hypotheticals related to the case or mention a car (until one of the prospective jurors brought it up themself).

3. Closing Argument

¶ 35 Gibbens-Martinez argues that the prosecutor committed misconduct during his closing argument by stating his personal

19

opinion about and improperly commenting on the defendant’s credibility. Again, we perceive no plain error.

a. Additional Facts

¶ 36 As discussed, Gibbens-Martinez’s theory of defense at trial centered around whether she knew that she was driving a stolen car. Although she didn’t testify in her own defense, the statements she made to Officer Faivre about not knowing or believing that the car was stolen were admitted at trial.

¶ 37 In his closing argument, the prosecutor told the jury,

[E]very single one of you, including the defendant, know[s] that it’s stolen. And just because you look at [the person she obtained the car from] and you say, “Hey. You didn’t steal this, did you?” And they say, “No. I’m just selling it out of the back of my car with the security tags still on it. It’s a Louis Vuitton purse, but I didn’t steal it.” That is not an excuse. That is not a defense. That is ridiculous. Okay. Now, that’s even if you believe [Gibbens-Martinez’s] story. This is not the most credible person in the world. Let’s keep in mind what situation she is in. She’s under arrest for being in possession of a motor vehicle that is stolen. That’s when she is telling a story about [the person she obtained the car from]. This is someone who said, “I didn’t even notice the front door was damaged.” The front door that you can’t open.

So I will submit to you that I don’t think that’s credible at all.

20

(Emphasis added.)

b. Analysis

¶ 38 Gibbens-Martinez argues that the prosecutor committed reversible misconduct because he stated his personal opinion on Gibbens-Martinez’s credibility and defense, stating that she wasn’t the “most credible person in the world” and calling her defense “ridiculous.”

¶ 39 “[E]xpressions of personal belief as to the guilt of the defendant by the prosecutor are improper.” Domingo-Gomez v. People, 125 P.3d 1043, 1049 (Colo. 2005). Indeed, a prosecutor may not “throw onto the scales of credibility the weight of his own personal opinion.” Wilson v. People, 743 P.2d 415, 418 (Colo. 1987) (citation omitted). Moreover, it’s “categorically improper” for a prosecutor to assert their opinion that a defendant has lied or is a liar. Wend, 235 P.3d at 1096.

¶ 40 The Colorado Rules of Professional Conduct and the American Bar Association Standards also require that lawyers avoid statements of personal opinion. Colo. RPC 3.4(e); Crim. Just. Standards: Prosecution Function § 3-6.8(b) (A.B.A., 4th ed. 2017). Colo. RPC 3.4(e) requires that lawyers not “state a personal opinion

21

as to the justness of a cause, the credibility of a witness, . . . or the guilt or innocence of an accused.” Standard 3-6.8(b) of the American Bar Association Criminal Justice Standards for the Prosecution Function explains that a “prosecutor should not argue in terms of counsel’s personal opinion, and should not imply special or secret knowledge of the truth or of witness credibility.”

¶ 41 We conclude that although the prosecutor’s comments were improper, they weren’t so flagrant as to require reversal. We reach this conclusion for three reasons. First, while the prosecutor commented on Gibbens-Martinez’s defense and said that he didn’t think someone who said that they didn’t notice the front door was damaged even though it couldn’t be opened was credible, the context of these arguments indicates that they were anchored in the evidence and testimony of witnesses. See Domingo-Gomez, 125 P.3d at 1051 (a reviewing court may consider “the context in which [a challenged] statement was made” to determine if it is improper); see also People v. Sanders, 2022 COA 47, ¶ 53 (“[A]lthough ill-advised, a prosecutor’s use of the first person singular does not automatically transform his expression of confidence into a personal opinion.”), aff’d on other grounds, 2024 CO 33.

22

¶ 42 Second, the comments on Gibbens-Martinez’s credibility or the believability of her defense didn’t suggest that some additional, unadmitted evidence supported her guilt or her counsel’s assessment of her guilt. See People v. Garcia, 2023 COA 58, ¶ 59. Instead, the comments were tied to the evidence admitted during trial.

¶ 43 Third, considering the brief comments in the context of the prosecutor’s closing argument as a whole, they didn’t so undermine the fundamental fairness of the trial as to cast serious doubt on the reliability of the verdict. See People v. Ujaama, 2012 COA 36, ¶ 74; see also People v. Rhea, 2014 COA 60, ¶ 71 (holding that the two challenged comments made by the prosecutor were “brief and isolated” and thus weren’t plain error); People v. Marko, 2015 COA 139, ¶ 223 (holding that the multiple challenged comments by the prosecutor weren’t plain error where they were “a small part” of the closing argument), aff’d on other grounds, 2018 CO 97.

C. Reasonable Doubt Instruction

¶ 44 Next, Gibbens-Martinez contends that the trial court reversibly erred by using the definition of reasonable doubt from the 2022 model criminal jury instruction because some of the language

23

omitted and included in that instruction lowered the prosecution’s burden of proof, thereby violating her due process rights. We disagree that reversal is warranted.

1. Additional Facts

¶ 45 Trial in this case took place in April 2023. A few months prior, on December 9, 2022, the Colorado Supreme Court Model Criminal Jury Instructions Committee promulgated a new model instruction defining reasonable doubt. See COLJI-Crim. E:03 (2022); Colo. Jud. Branch, Model Crim. Jury Instructions Comm., Meeting Minutes of: Friday, December 9, 2022, https://perma.cc/6WXQ- 29Y8. We will refer to this as the 2022 model instruction.

¶ 46 Notwithstanding the newly promulgated 2022 model instruction, Gibbens-Martinez asked the trial court to give the COLJI-Crim. E:03 (2021) instruction on reasonable doubt, which defines reasonable doubt as follows:

Reasonable doubt means a doubt based upon reason and common sense which arises from a fair and rational consideration of all of the evidence, or the lack of evidence, in the case.

It is a doubt which is not a vague, speculative or imaginary doubt, but such a doubt as would cause reasonable people to hesitate to act in matters of importance to themselves.

24

(Emphasis added.)

¶ 47 The court denied Gibbens-Martinez’s request and gave the 2022 model instruction instead. The instruction given at trial read, in pertinent part, as follows:

The burden of proof in this case is upon the prosecution. The prosecution must prove to the satisfaction of the jury beyond a reasonable doubt the existence of each and every element necessary to constitute the crime charged. This burden requires more than proof that something is highly probable, but it does not require proof with absolute certainty.

Proof beyond a reasonable doubt is proof that leaves you firmly convinced of the defendant’s guilt. If you are firmly convinced of the defendant’s guilt, then the prosecution has proven the crime charged beyond a reasonable doubt. But if you think there is a real possibility that the defendant is not guilty, then the prosecution has failed to prove the crime charged beyond a reasonable doubt.

If you find from the evidence that each and every element of a crime has been proven beyond a reasonable doubt, you should find the defendant guilty of that crime. If you find from the evidence that the prosecution has failed to prove any one or more of the elements of a crime beyond a reasonable doubt, you

25

should find the defendant not guilty of that crime.[1]

(Emphasis added.)

2. Standard of Review and Applicable Law

¶ 48 We “review jury instructions de novo to determine whether the instructions accurately informed the jury of the governing law,” Riley v. People, 266 P.3d 1089, 1092 (Colo. 2011), and a trial court’s decision to give, or not give, a particular jury instruction for an abuse of discretion, People v. Melara, 2025 COA 48, ¶ 16.

¶ 49 “Due process guarantees to the criminal defendant that the prosecution must prove every factual element necessary to constitute the crime charged beyond a reasonable doubt before the defendant may be convicted and subjected to punishment.” Vega v. People, 893 P.2d 107, 111 (Colo. 1995); see U.S. Const. amend.

1 After the trial in this case, the Colorado Supreme Court Model

Criminal Jury Instructions Committee again updated the reasonable doubt instruction, this time reinserting the reference to the “lack of evidence” by adding a new sentence that reads, “A reasonable doubt can be based on the evidence presented or the lack of evidence presented.” COLJI-Crim. E:03 (2023) (emphasis added); see also People v. Schlehuber, 2025 COA 50, ¶ 18 & n.2 (explaining that this change was made “without explanation”); People v. Melara, 2025 COA 48, ¶ 15 (“The committee provided no rationale for the additional sentence.”).

26

XIV, § 1; Colo. Const. art. 2, § 25. While trial courts are required to properly instruct the jury on the reasonable doubt standard, they have flexibility in defining it. Johnson v. People, 2019 CO 17, ¶ 10. But “[i]nstructions that lower the prosecution’s burden of proof below the reasonable doubt standard constitute structural error and require automatic reversal.” Tibbels v. People, 2022 CO 1,

¶ 22. To determine if a trial court’s instruction lowered the prosecution’s burden of proof, we look at the record as a whole and “ask whether there is a reasonable likelihood the jury applied the instructions in an unconstitutional manner.” People v. Garcia, 2021 COA 80, ¶ 26, aff’d, 2023 CO 30.

¶ 50 Model jury instructions are intended to be guidelines for trial courts to consider. People v. Morales, 2014 COA 129, ¶ 42. They aren’t, however, binding law, People v. Randolph, 2023 COA 7M, ¶ 22, aff’d, 2025 CO 44, nor are they “a safe harbor that insulates instructional error from reversal,” Garcia v. People, 2019 CO 64,

¶ 22. We must review the use of a model jury instruction just as we would any other jury instruction. See id. at ¶ 23.

27

3. Analysis

¶ 51 Gibbens-Martinez argues that the trial court impermissibly lowered the prosecution’s burden of proof by instructing the jury using the 2022 model instruction. She takes specific issue with (1) the omission of language telling the jury to consider the “lack of evidence” and (2) the inclusion of the “real possibility” language. She contends that both deviations from the 2021 model instruction impermissibly lowered the prosecution’s burden of proof. We conclude that the trial court’s use of the 2022 model instruction isn’t grounds for reversal.

a. Lack of Evidence

¶ 52 Prior divisions of this court addressed the omission of the “lack of evidence” language from the 2022 model instruction. In those cases, the divisions concluded that the omission doesn’t require reversal. See People v. Simms, 2026 COA 51M, ¶ 33; People v. Berumen, 2025 COA 93, ¶ 33; People v. Schlehuber, 2025 COA 50, ¶ 25; Melara, ¶ 32. While we agree with these divisions that when instructing a jury on the beyond a reasonable doubt standard, it’s best practice to instruct the jury to consider the lack of evidence in the case, see Simms, ¶ 24; Berumen, ¶ 32;

28

Schlehuber, ¶ 20; Melara, ¶ 24, we also agree that the omission of that language doesn’t lower the prosecution’s burden of proof, see Simms, ¶ 24; Berumen, ¶ 33; Schlehuber, ¶ 25; Melara, ¶ 32; see also Melara, ¶ 125 (Welling, J., specially concurring).

¶ 53 We agree with the division in Schlehuber that “the concept of reasonable doubt inherently invites jurors to consider what evidence is missing.” Schlehuber, ¶ 21. Moreover, the omission of an explicit instruction telling the jury it can consider the “lack of evidence” when reaching its verdict “is not tantamount to a prohibition on doing so.” Melara, ¶ 32; see Simms, ¶ 24; Schlehuber, ¶ 21; Berumen, ¶ 33. And the 2022 model instruction, read as a whole, bolsters that conclusion by informing the jury that (1) the prosecution bears the burden of proof; (2) the defendant is presumed innocent unless the prosecution has met its burden; and (3) the jury must consider “all the evidence” in making its final determination. Schlehuber, ¶ 22. Accordingly, we conclude that the omission of the “lack of evidence” language didn’t lower the prosecution’s burden of proof, so reversal isn’t warranted.

29

b. Real Possibility

¶ 54 Gibbens-Martinez also contends that the “real possibility” language in the 2022 model instruction impermissibly lowered the prosecution’s burden of proof. Again, four prior divisions of this court addressed this issue and concluded that the language is an accurate statement of the law. See Simms, ¶ 28; Berumen, ¶ 29; Schlehuber, ¶¶ 29-34; Melara, ¶ 30. This conclusion also finds support from federal courts that have considered the issue. See Schlehuber, ¶ 30 (collecting cases). We agree with these courts.

¶ 55 In Schlehuber, the division explained that “the phrase ‘real possibility’ correctly directs the jury not to acquit the defendant simply because it can conceive of some fanciful possibility that the defendant is not guilty.” Id. at ¶ 31. Indeed, “[n]othing in that phrase suggests that the defendant must be the source of the real possibility.” Id. at ¶ 34 (citation modified). And like in Schlehuber, the trial court here repeatedly instructed the jury that the prosecution bears the burden of proving each element of the crime beyond a reasonable doubt, so the real possibility language “simply explained the threshold the prosecution must overcome to do so.” Id. The instruction given at trial here stated, “The burden of proof

30

in this case is upon the prosecution.” And during voir dire, the trial court told the jury venire several times that the burden belonged to the prosecution and that the defendant doesn’t have the burden. Accordingly, we conclude that the “real possibility” language didn’t lower the prosecution’s burden of proof.

D. Cumulative Error

¶ 56 Last, Gibbens-Martinez contends that the cumulative effect of the errors she asserts on appeal requires reversal. “For reversal to occur based on cumulative error, a reviewing court must identify multiple errors that collectively prejudice the substantial rights of the defendant, even if any single error does not.” Howard-Walker v. People, 2019 CO 69, ¶ 25. “Stated simply, cumulative error involves cumulative prejudice.” Id.

¶ 57 Between the errors we found and those we assumed, we’ve identified three potential errors: the admission of the KBB report under the Confrontation Clause and the two instances of prosecutorial misconduct. But any error regarding the KBB report only went to the valuation of the car, not to whether Gibbens- Martinez stole the car or her theory of defense. And for the same reasons that the errors during voir dire and closing argument didn’t

31

individually warrant reversal, these errors when considered together didn’t deprive Gibbens-Martinez of a fair trial nor do they undermine our confidence in the verdict. See People v. Herdman, 2012 COA 89, ¶ 79 (concluding that the combination of two unrelated errors didn’t deprive the defendant of a fair trial); see also People v. Mendenhall, 2015 COA 107M, ¶ 82 (explaining that to warrant reversal, “the cumulative effect of the errors [must have] substantially prejudiced [the] defendant’s right to a fair trial”).

III. Disposition

¶ 58 The judgment is affirmed.

JUDGE BERNARD concurs.

JUDGE LUM concurs in part and dissents in part.

32

JUDGE LUM, concurring in part and dissenting in part.

¶ 59 I disagree with the majority’s conclusion that Officer Faivre laid adequate foundation to establish the authenticity of the Kelley Blue Book (KBB) report. I would therefore conclude that the trial court erred by admitting the report — and Officer Faivre’s testimony in which he read the value of the car off the report — into evidence. I would also conclude that the error wasn’t harmless and would reverse Gibbens-Martinez’s conviction for second degree aggravated motor vehicle theft as a class 5 felony. Because I agree with the remainder of the majority opinion, I would give the prosecution the option to retry the case or accept a conviction on second degree aggravated motor vehicle theft as a class 6 felony.

¶ 60 Therefore, I respectfully concur in part and dissent in part.

I. Authentication

¶ 61 As the majority correctly points out, the People needed to show that the document was what it purported to be — a KBB report estimating the value of a vehicle of a particular make, model, and year. See CRE 901(a). However, when the prosecutor handed the document to Officer Faivre and asked what it was, he testified only that it “appears to be” a KBB report. During the bench conference,

33

the prosecutor clarified that Officer Faivre had not prepared the supposed report, and the record doesn’t reflect that Officer Faivre had ever seen the document before it was handed to him on the witness stand. The majority concludes that the document was authenticated by Officer Faivre’s testimony that the make, model and year of the car on the document matched the car he found Gibbens-Martinez in. But while I agree with the majority that that testimony supported the document’s relevance, it did nothing to show that the document was what it purported to be.

¶ 62 While I acknowledge that the bar for authentication is low, it is not, in my view, this low. If the majority’s opinion is the rule, then all that is required to authenticate a document is for a witness who has apparently never seen it before to testify about what it “appears to be” based on nothing more than reading the document itself. That effectively makes any document self-authenticating, which is particularly troubling in an age when AI-generated evidence has created unprecedented authentication challenges for courts.1 See Connor Heaton, Shay Cleary & Michael Navin, AI-Generated

1 I do not suggest that the KBB report was AI-generated.

34

Evidence Is a Threat to Public Trust in the Courts, Nat’l Ctr. for State Cts. (Feb. 24, 2026), https://perma.cc/HY66-VP33.

¶ 63 Accordingly, I would conclude that the trial court abused its discretion by admitting the report and Officer Faivre’s testimony about it.

II. Standard of Reversal

¶ 64 The parties dispute whether this issue is preserved for review. Although a close call, under these circumstances, I conclude that it is because the trial court raised the authentication issue sua sponte, asked several questions about who prepared the report, and then impliedly determined that the report satisfied the authentication requirement of CRE 901 when it allowed the document into evidence. See Villanueva v. People, 199 P.3d 1228, 1232 (Colo. 2008) (although counsel’s argument before the trial court was vague, “the court was not only aware of the restrictions” before it but “actually considered these restrictions and nonetheless found the admission to be proper”); United States v. Hernandez- Rodriguez, 352 F.3d 1325, 1328 (10th Cir. 2003).

¶ 65 Preserved evidentiary errors are reviewed for harmlessness. Hagos v. People, 2012 CO 63, ¶ 12. Under this standard, we will

35

reverse if the error “substantially influenced the verdict or affected the fairness of the trial proceedings.” Id. (citation omitted).

III. The Error Is Not Harmless

¶ 66 The jury convicted Gibbens-Martinez of second degree aggravated motor vehicle theft as a class 5 felony. Under the then- applicable version of the statute, second degree aggravated motor vehicle theft is a class 5 felony if the value of the vehicle is greater than $20,000. § 18-4-409(4)(a), C.R.S. 2022. The KBB report put the vehicle’s value at $23,526. The jury also heard evidence that (1) the victim purchased the vehicle for $56,000 about four years before Gibbens-Martinez was found in possession of the vehicle, and (2) the vehicle had damage to the driver’s side door handle and the ignition.

¶ 67 Without the KBB report, it would have been difficult for the jury to discern that the vehicle’s value was greater than $20,000, given (1) the length of time between the purchase and the crime and (2) the vehicle’s damage when it came into Gibbens-Martinez’s possession. (Notably, the People didn’t argue that Gibbens- Martinez stole the vehicle in the first instance. Instead, they argued that she knew it was stolen when she took possession of it because

36

of the ignition damage.) The report no doubt weighed heavily in the jury’s assessment of value under those circumstances. Therefore, I cannot conclude that the admission of the KBB report was harmless, and I would reverse the conviction.

¶ 68 Because I agree with the remainder of the majority opinion, I would remand the case for the prosecution to choose between conducting a new trial, see People v. Marciano, 2014 COA 92M-2, ¶ 46, or having the trial court enter a judgment of conviction and resentence Gibbens-Martinez for second degree aggravated motor vehicle theft as a class 6 felony (value of the vehicle between $2,000 and $20,000), see § 18-4-409(4)(b).