When Technology Testifies: ShotSpotter, Due Process, and the Limits of Sound Technology
By Samantha Mazzone
As part of their recent decision in Commonwealth v. Rios, 496 Mass. 11 (2025), the Supreme Judicial Court (“SJC”) addressed evidentiary concerns involving a technological tool currently at the forefront of the law enforcement investigative arsenal: ShotSpotter. Specifically, the SJC denied the defendant’s motion for a new trial under the claims of ineffective assistance of counsel and newly discovered evidence, both arising from the Commonwealth’s use of ShotSpotter evidence at trial. Rios and other cases nationwide provide useful guidance on the factors judges are applying to decide if, when, and how ShotSpotter information may be introduced at trial.
What is ShotSpotter?
ShotSpotter is an acoustic gunshot detection technology used by law enforcement to identify, locate, and respond to shootings. Microphones inside black boxes attached to telephone poles capture gunshot decibels, triangulate their location, and send the data to an acoustic expert employed by ShotSpotter’s parent company, SoundThinking. The acoustic expert interprets the detected “gunshots” alongside their “sophisticated machine algorithms” to “ensure and confirm that the events are indeed gunfire.” See ShotSpotter Frequently Asked Questions. SoundThinking’s acoustic experts can “append the alert with other critical intelligence such as whether a fully automatic weapon was fired or whether there are multiple shooters,” all in sixty seconds. Id. The captured sounds and expert analysis are delivered to law enforcement in the same timeframe. Id.
A detailed forensic report is later provided containing both computer-generated and human expert analysis, though the report does not distinguish between the two. Some reports include an initial disclaimer notifying the reader of human intervention with the data. However, the report does not typically specify where the data has been modified, leaving questions about the amount of discretion acoustic experts have in interpreting the raw data on a case-by-case basis and the reliability of the reports. Although SoundThinking discloses some details regarding their specialized software in their annual reports, they “rely in part on trade secrets, proprietary know-how and other confidential information to maintain [their] competitive position.” See SoundThinking 2022 Annual Report at 46.
Commonwealth v. Rios
A Springfield jury found Lee Manuel Rios guilty of murdering Kenneth Lopez. Rios, 496 Mass. at 12. After discovering Lopez’s body, the police determined he died by multiple gunshot wounds and recovered the relevant ShotSpotter data from the closest activation site. Id. at 13, 18. The police used SoundThinking’s automatically generated summary report to support a finding that five shots were fired approximately two hundred to three hundred yards from where they found Lopez’s body. Upon police request, a SoundThinking employee provided a detailed forensic report that mapped the same five gunshots to the exact Dwight Street address where the police found Lopez’s body. The SoundThinking employee informed police that he determined the more precise location based on his manual recalibration of the data as part of his “postprocessing” analysis. Id. at 18-19.
The Commonwealth agreed to provide all data, report information, and communications exchanged with SoundThinking, but objected to providing the mechanics behind ShotSpotter’s location calculation and calibration process. Id. at 21. Before trial, Rios filed a motion to preclude the SoundThinking expert’s testimony because SoundThinking had “refused to share, or allow (the expert) to share, the underlying ‘mechanics,’ ‘calculations,’ ‘algorithms,’ and ‘data’ of the manual recalibration process” or “sufficient information to enable the defendant to review and test its calculations.” Id. at 25. The trial judge denied the defendant’s motion and the expert testified to the general mechanics of ShotSpotter and the postprocessing analysis he conducted.
After the jury found Rios guilty, he filed a motion for new trial arguing, inter alia, (i) that his counsel was ineffective because she failed to hire an audio expert to rebut the reliability of the ShotSpotter results and (ii) that testimony from a SoundThinking employee in a California criminal case constituted newly discovered evidence that undermined the reliability of ShotSpotter’s postprocessing methods. Id. at 24. The trial judge denied the motion and the SJC reviewed Rios’ conviction pursuant to its authority under G.L. c. 278 § 33E. Id. at 39.
SJC’s Ruling and Reasoning
To support his claim of ineffective assistance before the trial court, Rios submitted an affidavit and report from an electrical engineer that attested the ShotSpotter forensic report lacked sufficient detail “to assess the accuracy of ShotSpotter’s calculations.” Id. at 24. To support his claim of newly-discovered evidence, Rios submitted testimony of a SoundThinking manager in a California criminal trial who testified that analysts manually modify the ShotSpotter data, making the adjustments subjective, and that neither the postprocessing functions nor the system’s overall gunshot detection capabilities had been validated by outside sources. Id. at 27.
In holding that Rios’ trial counsel was not ineffective, the SJC pointed to counsel’s vigorous cross-examination of the Commonwealth’s expert witness and the testimony elicited challenging ShotSpotter’s reliability. The expert identified deficiencies in ShotSpotter’s overall accuracy and disclosed he had “undocumented discussions” with Springfield police about the case before he finalized the “ostensibly objective geolocation” of the shooting. Id. at 25. He also acknowledged SoundThinking’s guarantee of ShotSpotter’s ability to accurately capture “eighty five percent of all detectable events” within a specified radius was merely a general policy statement and not actually a guarantee. Id. Finally, the expert admitted essential sounds and echoes critical to his postprocessing analysis were not contained in his forensic report, and that he had “‘manually selected some of the data and reprocessed it through a process that [he] won’t disclose.’” Id. The SJC held that trial counsel’s vigorous cross-examination effectively highlighted ShotSpotter’s reliability issues to the jury and that additional expert opinion would have been cumulative and would not have meaningfully influenced the jury’s determination of Rios’ guilt. Id.
In holding that no new trial was warranted, the SJC concluded that, regardless of the ShotSpotter analysis, the prosecution had provided sufficient evidence to establish both the time and location of the shooting and evidence of Rios’ guilt independent of the ShotSpotter data. Id. at 27-28. Specifically, not only did a witness record Rios saying he killed the victim by shooting him five times and “‘count[ed] the houses’” to indicate his location, he also wrote a letter to his sister stating, “‘[T]hat’s why I did it by myself.’” Id. at 18, 20. The letter placed Rios on Dwight Street just before the shooting, and he admitted to shooting a .38 caliber revolver, which matched the ammunition recovered from the scene and the victim’s body. Id. at 28. Therefore, any impeachment of the Shotspotter evidence would not “have been a ‘real factor in the jury’s deliberations’” where other evidence sufficiently established the time and location of the murder, number of shots fired, and “overwhelming evidence of the defendant’s guilt.” Id. at 28.
Judicial Treatment of ShotSpotter Evidence Nationwide
Judges around the country are grappling with applying the law of digital evidence to ShotSpotter technology. The Massachusetts standard jury instruction for authenticating digital evidence requires the prosecution to authenticate digital evidence beyond a reasonable doubt if such evidence is submitted as proof of an element of any offense. . If it is being submitted for any other purpose, such as corroboration of other testimony, the judge may admit the evidence if the authenticity is proven by a preponderance of the evidence. Id. Courts nationwide are applying similar principles in admitting or excluding ShotSpotter evidence.
In People v. Hardy, 65 Cal. App. 5th 312 (Cal. Ct. App. 2021) the California Court of Appeals reversed and remanded the defendant’s semi-automatic weapon conviction based on the trial court’s erroneous admission of ShotSpotter evidence to prove the type of weapon the defendant used in the shooting. Before trial, the court preliminarily indicated the ShotSpotter recording would be admitted for the limited purpose of corroborating the officers’ testimony. Id. at 321-22. However, the trial court then admitted the evidence at trial “without further explanation” despite the defendant’s objection. Id. The officers who witnessed the shooting could not identify the type of gun the defendant allegedly used but testified they heard six “and possibly seven” shots in a short period of time, which led them to believe he had fired a semi-automatic weapon. Id. at 317-18, 320. Consequently, the prosecution relied heavily on the ShotSpotter audio recording purporting to capture “seven distinct sounds, one right after another” as proof that the defendant possessed a semi-automatic weapon. Id. at 319.
The California Court of Appeals reversed the conviction, citing other cases where experts had testified about ShotSpotter’s margin of error, training protocols, and outdated mathematical models – all evidence missing in this trial. Where the ShotSpotter audio recording served as the “strongest and only unambiguous evidence” that Hardy fired a semi-automatic weapon, reversal was warranted. Id. at 327-28.
In J.A.R. v. State, 374 So. 3d 25 (Fla. Dist. Ct. App. 2023) the prosecution properly presented expert testimony to support its use of ShotSpotter evidence at trial. After ShotSpotter alerted officers to gunfire, the officers used real-time surveillance cameras to locate the defendant and found a firearm and magazine in his pocket. J.A.R., 374 So. 3d at 27. Upon appeal from the defendant’s conviction of multiple firearm charges, the J.A.R. Court found the trial judge properly admitted the recording and supportive testimony where the ShotSpotter expert, who had seven years of experience as SoundThinking’s forensic services manager and had testified about the technology 70 times, provided sufficient support for admission. Id. at 30.
In contrast, the prosecution in State v. Thornton, 309 So. 3d 366 (La. Ct. App. 2020) called a District Attorney Investigator, who was not an expert, to explain how ShotSpotter was used and accessed during the investigation and how the technology corroborated other evidence. Id. at 371-72. Where the prosecution did not use the ShotSpotter evidence as direct proof of an element, the Thornton Court found it properly admitted with the lay testimony. Id. at 372.
Finally, the Supreme Court of Pennsylvania joined the Massachusetts SJC on the short list of appellate courts that have used hearsay rules to evaluate under what circumstances the ShotSpotter forensic summary report may be admissible. In Commonwealth v. Weeden, 304 A.3d 333 (Pa. 2023), the Court upheld the trial court’s decision to admit the ShotSpotter forensic summary report to establish the time and location of the shooting and number of shots fired. The trial judge allowed a detective to testify about ShotSpotter’s mechanics and its role in the investigation. During cross-examination, the detective explained that after the computer generates the summary, ShotSpotter analysts may add updates based on information from responding officers, and admitted to not having ShotSpotter training, not personally working on the report, and believing that ShotSpotter was “not completely foolproof.” Id. at 336. Despite the defendant’s Confrontation Clause argument—where he claimed the summary report should have been excluded as testimonial rather than admitted as machine-generated because ShotSpotter analysts manipulated the data—the Court deemed the ShotSpotter analyst’s manipulation insufficient to categorize the document as human-created and, therefore, testimonial. Id. at 338.
Notably, in his concurrence, Justice Wecht scrutinized the disclaimer on the ShotSpotter summary report that disclosed the ShotSpotter analyst’s interpretation and work on the final product, stating it “bears many of the hallmarks [of] the type of unreliable evidence for which the Confrontation Clause is designed to test.” Id. at 359. Pointing to the detective’s admissions of ShotSpotter’s flaws, the vague disclaimer about human manipulation of critical data, and the company’s own warnings that the results should be “used with caution,” the concurrence expressed concern that the trial court admitted the report without requiring the government to present expert testimony from the report’s creator. The concurrence specifically took issue with the six-month gap between the date of the shooting and the date on which the summary report was created, noting that “there is at least a colorable argument that [the summary report] was not generated in response to an ongoing emergency.” Id. at 358. Despite the concerns and disagreement with the majority, however, the concurrence did not believe there was a legal basis to reverse the conviction based on Supreme Court precedent allowing non-testimonial statements to be admitted without confrontation. Id. at 362.
Conclusion
While the treatment of ShotSpotter evidence in Massachusetts is still evolving, the Rios decision aligns with the nationwide trend and Massachusetts digital evidence jury instruction for determining if digital evidence is admissible. Under this framework, the trial court should evaluate if the Commonwealth is offering the evidence (i) to corroborate other evidence of the crime, or (ii) to prove an element or an essential part of its burden. If the Commonwealth is using the evidence to corroborate other evidence, the burden of proof remains a preponderance of the evidence. If the Commonwealth is attempting to use digital evidence to prove an element of the crime or an essential part of its burden, the standard is raised to beyond a reasonable doubt. As exemplified in Rios, when ShotSpotter evidence is used only to support other independent proof—rather than to establish an essential element of the offense—the court is more willing to admit it, since the reliability of the conviction does not hinge solely on the accuracy of the digital evidence.
As it concerns ShotSpotter evidence, the proponent of the evidence should consider if support from a lay or expert witness is needed. Courts across the Commonwealth may look to factors similar to those considered for digital evidence generally to determine the reliability of ShotSpotter evidence in their own courtrooms, including reports assessing the reliability of the data and modifications made to the data by a SoundThinking acoustic analyst. ShotSpotter is a powerful technology that helps emergency responders save lives and attorneys facilitate justice. With the right framework, courts may properly evaluate the circumstances in which ShotSpotter data may be admitted as trustworthy evidence of proof at trial.
Samantha Mazzone is a third-year law student at Suffolk University Law School, a Paralegal at the Suffolk County District Attorney’s Office, and SJC 3:03 practitioner in the Suffolk Prosecutor Program.
