(rentas-nmcca-providence-ineffective-assistance-military-defense-lawyer)

United States v. Rentas: When “Trash” Almost Sank a Plea—And the IAC Claim That Followed

On 30 April 2026, the Navy-Marine Corps Court of Criminal Appeals (NMCCA) issued a per curiam opinion affirming the convictions and sentence in United States v. Rentas, No. 202500082. Lance Corporal Joshua J. Rentas had pleaded guilty at a general court-martial at Camp Foster, Okinawa, to four specifications of possessing child pornography and one specification of distributing child pornography under Article 134, UCMJ. The military judge sentenced him to a dishonorable discharge, 36 months of confinement, total forfeitures, and reduction to E-1. On appeal, Rentas raised two Grostefon issues: improvident pleas based on alleged pressure from counsel during a mid-providence recess, and ineffective assistance of counsel for failing to retain a digital forensic expert. NMCCA rejected both. The opinion offers a textbook walk-through of how providence inquiries work, what an “inconsistency” actually means, and how Strickland prejudice plays out in the guilty-plea context. For service members and any military defense lawyer working a pretrial agreement, it is worth a careful read.

The Facts the Court Cared About

Rentas confessed to NCIS, consented to the search and seizure of his electronic devices, and submitted a “letter of apology” with his proposed plea agreement. Pursuant to that agreement, he pleaded guilty to possessing child pornography on two laptops, an iPhone, and an external hard drive, and to one distribution specification.

During the providence inquiry on Specification 2 (the external hard drive), Rentas told the military judge he had put the hard drive in “the trash can.” Trial counsel called for a recess. Eight minutes later the inquiry resumed. Rentas explained “trash” actually meant a desk drawer “full of a bunch of trash,” that he had transferred the data—including some of the child pornography files—from the hard drive onto a different laptop, and that he planned to eventually get rid of the hard drive. The military judge accepted the pleas and imposed the agreed-upon sentence.

After trial, Rentas filed a declaration alleging that during the recess his trial defense counsel and the trial counsel both told him to “stick to the script” or his plea deal would collapse and he would face “over 50 years of confinement.” NMCCA initially denied his motion to attach the declaration, then denied reconsideration, then on its own reconsidered again and granted the motion. Even with the declaration in the record, the appeal failed.

Issue One: The Providence Inquiry and the “Trash” Problem

The standard of review for accepting a guilty plea is abuse of discretion, with questions of law reviewed de novo. United States v. Inabinette, 66 M.J. 320, 322 (C.A.A.F. 2008). Under Article 45, UCMJ, and R.C.M. 910(h)(2), the military judge must reject a plea—or resolve the inconsistency—if the accused sets up matter inconsistent with the plea. The substantial basis test asks whether anything in the record raises a substantial question about the plea.

NMCCA held nothing in the record raised that question for Specification 2. The “trash” comment was not an inconsistency. Rentas explained the desk-drawer-full-of-trash meaning, admitted he transferred the files including the child pornography to another device, and never claimed he was not guilty or that he obtained the files inadvertently. The stipulation of fact had him admitting “exclusive dominion and control” of the hard drive. That is the opposite of an inconsistency. The court distinguished United States v. Hayes, 70 M.J. 454 (C.A.A.F. 2012), and United States v. Phillippe, 63 M.J. 307 (C.A.A.F. 2006): a military judge must dig deeper when the accused raises a possible defense, but mere mitigation or rationalization is not the same thing as a defense. See United States v. Shaw, 64 M.J. 460 (C.A.A.F. 2007).

The recess-pressure allegation got even shorter treatment. Rentas did not argue coercion. The court flatly stated that “it is reasonable for a defense counsel to inform a client of possible and foreseeable consequences if a military judge were to reject a guilty plea as improvident.” Telling a client what the maximum punishment looks like if the deal collapses is not improper pressure—it is candid lawyering. The court also pointed out that Rentas did not challenge providence as to Specifications 1, 3, 4, or 5, but asked that all five findings and the sentence be set aside. NMCCA was not biting.

For background on how the Care inquiry functions and how plea problems develop in the record, see our prior post: Guilty pleas under the UCMJ.

Issue Two: Ineffective Assistance and the Digital Forensic Expert

Rentas’s second claim was that trial defense counsel was ineffective for failing to retain a digital forensic expert who could have explained “the process by which files can be transferred from one device to another without the knowledge or intent of the user.” The court reviewed the claim de novo under Strickland v. Washington, 466 U.S. 668 (1984), and the guilty-plea-specific prejudice standard from Lee v. United States, 582 U.S. 357 (2017): whether there is a reasonable probability that, but for counsel’s errors, the accused would not have pleaded guilty and would have insisted on going to trial.

The opinion did two things that matter for any future Strickland claim in the plea context. First, it presumed reasonable tactical judgment. Rentas had already confessed to NCIS and consented to the searches that produced the devices. Against that record, declining to retain a digital forensic expert was a reasonable strategic call—and there is “significant” evidence to overcome before that presumption breaks. Second, it applied Suarez and Furth. A bare post-trial assertion that the accused would not have pleaded guilty is not enough. The court must look at contemporaneous evidence—the confession, the consent search, the apology letter, the stipulation of fact, the agreed sentence cap—because every appellant has an incentive to retroactively reframe the deal. United States v. Suarez, 86 M.J. 65 (C.A.A.F. 2025); United States v. Furth, 81 M.J. 114 (C.A.A.F. 2021).

The court added a sharp factual finding tucked in a footnote: it did not credit Rentas’s claim that he was unaware files would download onto the hard drive. With that credibility call, the prejudice analysis was over.

What This Opinion Teaches

Several practical points emerge. A providence inquiry is not a verbatim test. The military judge must resolve actual inconsistencies, not chase every mitigating word. Rationalization, minimization, and clarification do not invalidate a plea unless they negate an element or set up a defense. See our discussion of this principle: Providence shines down or upon.

Recess conversations between an accused and defense counsel about the consequences of an improvident plea are normal and lawful. Allegations of “pressure” must show coercion that overrides voluntariness, not just candor about exposure if the deal collapses. A post-trial declaration alone, even if attached to the record, rarely moves the needle when the contemporaneous record—confession, consent, stipulation, apology letter, plea agreement—uniformly supports the plea.

Ineffective-assistance claims in the plea context turn on whether the accused would have rejected the deal and gone to trial. That is an objective inquiry, anchored to the actual evidence of guilt, not to the appellate brief’s hindsight. When the Government has a confession, consensual search results, and a stipulation, the Lee prejudice showing is steep.

Digital forensic experts can be powerful tools in CSAM and computer-evidence cases. See United States v. Douangdara and our prior post: How a guilty plea can go wrong. But timing matters. The right place to litigate digital forensics is before the guilty plea, not after. Once an accused confesses, consents to searches, signs a stipulation, and pleads guilty, the appellate burden to undo the plea on a forensics theory becomes very heavy.

The Strategic Lesson Before You Sign a Plea Agreement

If digital forensics could exonerate or mitigate—if there is a real argument about automatic syncing, cloud backups, malware downloads, shared devices, or unintended file transfers—those arguments belong in pretrial motions, suppression litigation, expert assistance requests under R.C.M. 703(d), and pretrial negotiations. They lose nearly all their power once the providence inquiry is over and the plea is accepted. Rentas is the latest reminder that the leverage in a CSAM case is built before arraignment, not after sentencing.

How Cave & Freeburg Can Help

At Cave & Freeburg, our military defense lawyers handle Article 134 child pornography cases, NCIS, OSI, CID, and CGIS investigations, and the full range of pretrial and appellate work in courts-martial across every service. We litigate motions to suppress under the Fourth and Fifth Amendments, secure expert assistance for digital forensics under R.C.M. 703(d), evaluate plea agreements with realistic exposure analysis, and—when the government’s case has real weaknesses—we contest at trial rather than settle. On appeal, we know how NMCCA, AFCCA, ACCA, the Coast Guard CCA, and CAAF handle providence challenges, Strickland claims under Suarez, Furth, and Lee, and post-trial declarations. If you face a child pornography investigation or charge under Article 134, UCMJ, or you have already been convicted and want a candid appellate assessment, contact Cave & Freeburg, LLP, today. Call (703) 298-9562 or (917) 701-8961 to speak confidentially with an experienced military defense counsel.

James J. Duane, reminds us of the need to pay attention to word choice in his article, “The Right to Remain Silent: A New Answer to an Old Question.”

There is no official language that a witness is required to employ when invoking the privilege against self-incrimination. As one federal circuit court recently observed, “A witness’s answer could range from ‘I refuse to answer on the ground that my answer may tend to incriminate me’ to the more mundane ‘On the advice of counsel, I decline to answer.’” Evans v. City of Chicago, 513 F.3d 735, 740 n.4 (7th Cir. 2008).

One point the author makes is that the usual language implies the person is guilty but refuses to admit it. While that may not be admissible in court (Duane offers examples where it could get to be evidence), it casts a pall over the investigation.

Duane suggests one option is:

“On the advice of my lawyer, I respectfully decline to answer on the basis of the Fifth Amendment, which—according to the United States Supreme Court—protects everyone, even innocent people, from the need to answer questions if the truth might be used to help create the misleading impression that they were somehow involved in a crime that they did not commit.”

He suggests the expanded language is better than just saying “I decline to answer based on my Fifth Amendment right to silence.” I would add that should you invocation become “evidence” either deliberately or accidentally, the broader language would go over better with the Members (jury).

Sundby on Interrogation Law

By CrimProf BlogEditor

Share Scott E. Sundby (University of Miami School of Law) has posted The Court and the Suspect: Human Frailty, the Calculating Criminal, and the Penitent in the Interrogation Room (Washington University Law Review, Vol. 98, No. 1, 2020) on SSRN. Here is the abstract: For the past century, the Supreme Court has swung back-and-forth between a laissez faire approach to police interrogation and the far more regulatory approach exemplified by Miranda v. Arizona. Indeed, it would be quite understandable if someone were to develop constitutional vertigo in tracing the Court’s meanderings through the various constitutional byways of the Due Process clause, the Sixth Amendment right to counsel, and the Fifth Amendment privilege against self-incrimination. The Court’s failure to bring coherence to its case law has offered up a legal piñata for scholars and judges to take a swing at as the search continues to figure out the proper role for the Bill of Rights in regulating police interrogation.

This Article brings a new perspective to that quest by arguing that much of the Court’s muddled jurisprudence regarding police interrogation is a result of the Justices’ differing views of why individuals confess. The linchpin to all of the Court’s various approaches to police interrogation is the deceptively simple question of whether the suspect “voluntarily” decided to confess and waive his rights. As it turns out, answering the voluntariness question requires making assumptions about human behavior that transform the decision maker into as much psychotherapist as fact finder. By focusing intently on who the Justices envision as the suspect being questioned and the reasons why they confess, we discover that the Justices over time have espoused two fundamentally different characterizations of who is in the interrogation room: that of the “rugged individual” and that of the “susceptible individual.” Given that these two distinct views of the human psyche reside side-by-side within the Court’s jurisprudence, it becomes understandable that the Court’s confession law has a distinctly schizophrenic cast.

This framework, however, does more than to help explain the confusion surrounding the case law. By bringing these two competing visions out into the open, a direct examination from both an empirical and doctrinal viewpoint can be made of their underlying assumptions about human behavior. This examination shows that much of the Court’s current reliance on the rugged individual view is premised on assumptions that are belied by science. And, as the Article explains, the choice of characterization not only has a profound impact on how the constitutional rule is shaped, but has critical ramifications for matters such as lost convictions, false confessions, and the balance of power between the government and citizen.

This is an interesting discussion from another country on something we are familiar with.

It will be recalled that the right to silence formerly comprises the privilege against self-incrimination and the right not to have adverse inferences drawn from his silence. Prior to the CJPOA, no evidential significance could be attached to an accused’s exercise of the right to silent, save when the accused and the victim were on even terms. However, theprovisions in the CJPOA, particularly ss 34-35 have sought to alter this principle to the extent that the question that this essay grapples with, is whether the right to silence,despite the changes, is still useful in protecting an accused’s supposed ‘constitutionalright’ of innocence, until proven guilty.

We are familiar with this–

Police Interview– a search for truth or guilt? To start with, current police interviewing guidelines describe the investigative interview as a ‘search for truth’. But this assumption to some academics is itself mired. As Kelly Benneworth-Gray argues, ‘truth’ and ‘honesty’ are locally invoked interactional resources, produced, recognised and contested in two very different sequential environments. The upshot of this is that regardless of the offer of truth made by the accused, the main purpose of a police interview is to mark a disjuncture between the testimonies of the suspect and the alleged victim and construct the suspect’s testimony as implausible.

Thus, Article 31(b), UCMJ, warnings are required when (1) a person subject to the UCMJ, (2) interrogates or requests any statement, (3) from an accused or person suspected of an offense, and (4) the statements regard the offense of which the person questioned is accused or suspected.” Jones, 73 M.J. at 361 (footnotes omitted) (citation omitted). However, the second of these prongs is met only if the questioner was acting in an official law enforcement or disciplinary capacity, or could reasonably be considered to be acting in such a capacity by a “reasonable person” in the suspect’s position. Id. at 362. “Questioning by a military superior in the immediate chain of command ‘will normally be presumed to be for disciplinary purposes,’” although such a presumption is not conclusive. Swift, 53 M.J. at 446 (quoting United States v. Good, 32 M.J. 105, 108 (C.M.A. 1991)) (additional citations omitted). An “interrogation” includes “any formal or informal questioning in which an incriminating response either is sought or is a reasonable consequence of such questioning.” Mil. R. Evid. 305(b)(2).

United States v. Ramesk, AFCCA, 29 Oct. 2018, at 17.

In United States v. Cox, decided by the Navy-Marine Corps Court of Criminal Appeals in November 2018, the Appellant raised three issues, one of which asked whether statements to a [civilian but military retiree] recovery care coordinator should be suppressed because they were not preceded with an Article 31, UCMJ, warning. The initial problem for Cox was her counsel’s failure to raise the issue prior to trial but waited until after entry of pleas. The CDC made the motion late–note this was a civilian counsel case (CDC). The CDC waited until testimony on the merits. It appears that the CDC acted “to gain a tactical advantage over the government.” Slip op. at 5. What advantage was to be gained is not clear from the record, other than an unprepared TC who now has to answer a significant motion on the fly.

(In practice, you should be careful when entering pleas. If possible delay entry of pleas as long as possible, most judges will allow a delay until a later date specified in the trial management order or will allow counsel to defer motions to a date set in the trial management order regardless of the entry of pleas.)

Here, the military judge did not find good cause for the delay in moving to suppress statements and found the issue was waived. Regardless, the military judge appears to have addressed the suppression issue (perhaps concerned that the NMCCA or CAAF might not find the issue waived). The military judge ruled that the civilian employee witness was not required to advise the accused of her right to silence. NMCCA found the issue was waived, but like the military judge addressed the suppression issue anyway and held the military judge did not abuse his discretion by admitting any incriminating statements.

In addressing the IAC claim for failure to raise the motion timely, NMCCA found there was no prejudice because the motion would have failed anyway. It will be interesting to see how CAAF if it grants a petition, deals with the issues. The law on civilian employees of the military and their obligation to give Article 31, UCMJ, warnings is reasonably clear at this point, but perhaps CAAF wants to put another nail in the coffin of a civilian employee’s Article 31, UCMJ, “obligation” and leave the question for resolution on a case-by-case basis. If the CAAF does find the statements should have been suppressed then they get to the waiver and IAC issue.

Of the 1,705 post-conviction DNA and non-DNA exonerations that have occurred from 1989 to the end of 2015, approximately 13 percent of these wrongful convictions were due to false confessions, and virtually all of these occurred in either homicide or rape cases. This chapter discusses why false confessions occur and discusses the ways that law enforcement training can be modified to avoid false confessions. False confessions primarily occur due to a lack of proper training, poor investigative practices, and the use of scientifically invalidated and/or high risk interrogation techniques and strategies. To safeguard against false confessions, the author argues that investigators should receive training on the following topics: 1) the existence, variety, causes and psychology of false confessions; 2) the indicia of reliable and unreliable statements and how to distinguish between them; 3) the need to obtain corroborating evidence to verify suspects’ confessions; and 4) avoidance of inadvertent contamination of interrogations by disclosure of non-public case facts to suspects.

Leo, Richard A., Interrogation and False Confessions in Rape Cases (December 2015). in Robert Hazelwood and Ann Burgess, eds., PRACTICAL ASPECTS OF RAPE INVESTIGATION: A MULTIDISCIPLINARY APPROACH (CRC Press, 5th ed., 2016 Forthcoming); Univ. of San Francisco Law Research Paper . Available at SSRN: http://ssrn.com/abstract=2700410

Some points for consideration when seeking to suppress a confession. General risk factors include:

• Length of the interrogation.
• False Evidence. Social science research has shown that false evidence ploys are virtually always present in interrogations leading to false confession, and are substantially likely to increase the risk of eliciting false confessions from innocent suspects.
• Minimization.
• Threats and Promises.
• Individuals who, by their nature and personality, are naive, excessively trusting of authority, highly suggestible and/or highly compliant and who are therefore predisposed to believe that they have no choice but to comply with the demands of authorities or who simply lack the psychological resources to resist the escalating pressures of accusatorial interrogation.
• Contamination is the leakage or disclosure to a suspect of non-public case facts that are not likely guessed by chance.

The notorious CDR (judge)(Art. 32 IO) Monahan has this piece.

A PROPOSAL TO CONFORM MILITARY RULE OF EVIDENCE 305 IN LIGHT OF THE SUPREME COURT’S HOLDINGS IN MARYLAND V. SHATZER AND BERGHUIS V. THOMPKINS

The Shatzer Court held that a fourteen-day break in custody will end the Edwards presumption that police-initiated custodial interrogation after a suspect invoked his right to counsel is involuntary. The Shatzer Court also held that when an interrogated suspect who is being held in incarceration due to a prior conviction is released back to the general prison population, this constitutes a break in custody with regard to the termination of the Edwards protection discussed above. Together, the holdings of the Shatzer case should be applied to revise MRE 305(e)(3)(A), which sets forth the military rule for interrogation of an accused or suspect who requests counsel and who is in custody, could reasonably believe himself or herself to be in custody, or is otherwise deprived of his or her freedom in any way.

(Note: this is the military Vaughters Rule. United States v. Vaughters, 44 M.J. 377, 378 (C.A.A.F. 1996) (custodial interrogation may be reinitiated without counsel being present where a suspect is provided a meaningful opportunity to consult with counsel, and subsequently waives his right to counsel). Note further that in Vaughters CMA said a 19 day lapse was fine.)

In Berghuis v. Thompkins, the Supreme Court held that a suspect who has properly received and understood the Miranda warnings waives the right to remain silent by making an uncoerced statement to the police without invoking the Miranda rights. Moreover, the Thompkins Court held that the police are not required to obtain a waiver from a suspect of his or her right to remain silent before commencing interrogation. Applying the holdings of the Thompkins case, MRE 305(c)(4), that states the military rule concerning the exercise of the privilege against self-incrimination and the right to counsel, as well as MRE 305(e), that provides the military rule regarding an accused’s or suspect’s waiver of the rights provided under MRE 301 and MRE 305, should be revised.

1. Spurling claimed IAC because his counsel did not litigate his admissions. Interestingly both counsel admitted they didn’t even catch the issue: [Counsel] failed to “recognize the issue based on [her] lack of experience, the work load at the time, and never having argued an Article 31 issue[.]”
2. Capt B concurs, stating that had the issue occurred to him “[he] would have proposed filing it.”
3. Both TDC acknowledge that it was not until after participating in a post-trial debrief with the military judge, who asked whether they had filed a suppression motion, that they recognized the issue.

So how is this admitted “failure” to be reviewed. The court states the standard as a need to show a reasonable probability the motion would be a success, and this must be a substantial chance, not a mere probability. United States v. Jameson, 65 M.J. 160 (C.A.A.F. 2007); United States v. McConnell, 55 M.J. 479, 482 (C.A.A.F. 2001). Without this finding, there is no IAC even if there is an error, because there is no prejudice. However, the dissenters, expressed through Judge King would set aside the findings and sentence on the IAC issue. The court then goes into the ongoing issue of when is a person acting within a official capacity. Finding no substantial likelihood of success on the motion the court finds no harmful and prejudicial error. There has been some discussion already on the requirement to advise a person of their Article 31, UCMJ, right to silence.

Expect to see Spurling at CAAF, potentially as a trailer.

Spurling wins something on the inappropriate sentence – a set aside of the BCD.

Do you have to raise every single motion? No.

Do you have to raise every single motion the client asks you to? No.

I wonder what the MJ would have done if the counsel had asked for a post-trial session to litigate the motion at that time, if for no other reason than to make a record for appeal? We know the MJ can hold such a hearing.

Yeah, right.

First they are told and usually escorted to the LE office. The escort won’t tell them why or what’s going on. They then have to wait the appropriate time in the waiting area to heighten the tension. I was reminded of this by a post from fourth amendment blog.

Continuing defendant’s open container stop for a beer can because he was nervous was unreasonable. “[L]ights and sirens at three o’clock in the morning could make a saint nervous without shedding any light at all on whether there was alcohol in the can.” United States v. Hemingway, 2013 U.S. Dist. LEXIS 34517 (D. D.C. March 13, 2013).*

I think the same applies to law enforcement interrogations. A LawFirms.com blog starts out:

Talking with police officers is usually one of the more stressful encounters we have in our lives, and one that we typically avoid at all costs. Even when we’ve committed no crime, it can be nerve racking, but when we’re guilty it’s much worse.

Of course there are other “indicia” offered as to why they disbelieved the accused. But, cautions those guru’s of interrogation (The Reid):

In conclusion, because laughter and humor relieve anxiety, it is common for both truthful and deceptive suspects to engage in these behaviors during an interview. The mere presence of laughter or attempted humor during an interview should not be considered a behavior symptom of deception.

The Reid also points out that:

Furthermore, when accused of wrong-doing, the tendency to deny opportunity, access, motive and propensity occurs within both innocent and guilty suspects.

You can read more about interrogations at The Reid. Why is it valuable for a defense counsel to read The Reid. Well of course it’s an aid in understanding how law enforcement may have coerced a confession, or got it wrong. But, it’s also an invaluable guide in how you interview your own client and witnesses. I’m not saying you become an investigator or accusatory toward your client or witnesses. The idea of interrogation techniques is to get information. I know law enforcement is only looking for the confession, but you have a broader purpose.

The Reid’s final caution today is:

Some research has attempted to identify specific cues associated uniquely with lying (nature of eye contact, micro tremors in the voice, unique facial expressions, etc.). These efforts have not produced accuracies much above chance levels. . . . Not all innocent or guilty suspects respond exactly the same way when questioned about a crime. . . . In conclusion, many laboratory studies investigating the validity of behavior symptom analysis are flawed because they attempt to identify specific behaviors that reveal truth or deception; indeed, there are no behaviors unique to truth or deception.

Of course that leads to my favorite law enforcement response – confirmation bias.

Gary L. Wells

Iowa State University, Department of Psychology

Elizabeth F. Loftus

University of California, Irvine – Department of Psychology and Social Behavior
January 16, 2013
Handbook of Psychology, Vol. 11, 2013, Forensic Psychology, Chapter 25, R.K. Otto and & I.B. Weiner (Eds), Hoboken, NJ: John Wiley & Sons, Inc.
UC Irvine School of Law Research Paper No. 2013-88

Abstract:
This chapter begins with a summary of the case of Thomas Brewster, who was tried for murder based in large part on eyewitness testimony. Ultimately DNA came to Brewster’s rescue, and he was freed before the trial ended. Analyses of taped interviews in the case help reveal how the interviewing process itself may have tainted the eyewitness testimony. The chapter continues with discussions of new psychological research on memory for complex events. This work shows how the details of events can be changed when witnesses are exposed to post-event information that is misleading. And with enough suggestion, entire events can be planted into the mind of ordinary healthy adults. The final section discusses new findings concerning eyewitness memory for people. This includes eyewitness identification of previously seen strangers, and new findings on procedures that can reduce mistaken identifications.

The petition of the day is:

New Mexico v. Herring

Issue: Whether Berghuis v. Thompkins requires advice that a suspect has the right to stop talking at any time in order to establish an implied waiver of Miranda rights.

This case illustrates, again if anything did, the value of video and audio recorded interrogations. Oooops. I meant suspect interviews. Interrogation is of course a loaded word, so investigators are cautioned to use the less accusatory and loaded terminology. Just more of the psychology of getting confessions.