ShotSpotter's "unreliable" audio was used as scientific evidence to accuse and convict a Black man of attempting to shoot Rochester's city police, whose conviction was later reversed by a county judge.
.avif){kind=small}
ShotSpotter's audio detection system was used as forensic evidence in the trial of a man accused of attempting to shoot a police officer. The system's analysis, which was subject to human review and modification, was later deemed unreliable by a judge, leading to the reversal of the defendant's weapons possession conviction. The case highlights concerns regarding the use of proprietary, non-peer-reviewed AI tools as primary evidence in criminal proceedings.
ShotSpotter's "unreliable" audio was used as scientific evidence to accuse and convict a Black man of attempting to shoot Rochester's city police, whose conviction was later reversed by a county judge.
The ShotSpotter system failed to reliably detect and classify gunshots, requiring subjective human modifications that made the forensic evidence scientifically unreliable for a criminal trial.
The incident arose post-deployment due to the unintentional failure of the ShotSpotter AI system's algorithms and forensic analysis process to provide reliable scientific evidence, leading to a wrongful conviction.
High Risk: The AI system is used in law enforcement and judicial contexts (specifically as forensic evidence in criminal trials), which has significant implications for safety, fundamental rights, and public interests under the EU AI Act's high-risk categories.
Severity ratings are based on this scale.
Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.
Reported: The report describes financial costs associated with the ShotSpotter contract and implies personal financial losses for the defendant.
Directly caused: The city of Rochester pays an annual contract fee of $130,000 to ShotSpotter, which includes trial testimony.
Indirectly caused: Silvon Neal likely suffered significant indirect financial losses, including lost wages and legal defense costs, during his 18-month wrongful incarceration.
Inferred additional harm: Other municipalities using ShotSpotter incur similar annual licensing fees, and other wrongfully accused individuals face substantial legal fees and lost income.
Reported: The report explicitly describes differential treatment of groups through the targeted deployment of the system.
Indirectly caused: The deployment of ShotSpotter sensors exclusively in lower-income and high-crime neighborhoods results in indirect differential treatment, disproportionately exposing minority residents to the risk of unreliable algorithmic surveillance and forensic errors.
Inferred additional harm: Widespread deployment of acoustic surveillance in marginalized communities across 90 cities likely perpetuates systemic bias and unequal policing outcomes.
Reported: The report explicitly describes violations of civil rights caused directly and indirectly by the incident.
Directly caused: The use of unreliable ShotSpotter evidence directly contributed to the wrongful conviction and 18-month incarceration of Silvon Neal, violating his civil rights.
Indirectly caused: The reliance on proprietary, unverified AI evidence undermines the constitutional right to a fair trial and the ability of defendants to confront the evidence against them.
Inferred additional harm: Numerous other defendants in the 90 cities using ShotSpotter may have had their civil rights compromised through the introduction of unscientific acoustic forensic evidence.
Reported: The report explicitly describes potential privacy risks associated with acoustic monitoring.
Indirectly caused: The continuous acoustic monitoring of public spaces by ShotSpotter sensors presents an indirect risk of capturing private street-level conversations without consent.
Inferred additional harm: The deployment of thousands of microphones in public areas across 90 cities creates a persistent, unquantified risk of mass surveillance and privacy violations.
Reported: The report explicitly describes psychological distress resulting from wrongful prosecution and incarceration.
Indirectly caused: Silvon Neal suffered significant psychological distress, anxiety, and trauma from being wrongfully incarcerated for 18 months and facing a potential life sentence based on unreliable AI evidence.
Inferred additional harm: Other defendants facing prosecution based on unreliable ShotSpotter evidence likely experience severe psychological distress and anxiety due to the threat of wrongful conviction.
Reported: The report explicitly describes epistemic harm through the generation of unreliable forensic evidence presented as scientific truth.
Directly caused: The ShotSpotter system and its analysts generated unreliable acoustic analyses, falsely presenting subjective human modifications of helicopter noise as scientific proof of five gunshots.
Indirectly caused: The introduction of unscientific algorithmic outputs in court briefs corrupts the judicial information environment and undermines the shared understanding of forensic truth.
Inferred additional harm: The widespread use of proprietary, non-peer-reviewed AI tools in criminal trials risks systemic epistemic corruption of the justice system.
The incident involved the use of unreliable ShotSpotter acoustic AI forensic evidence in a local criminal trial, leading to a temporary wrongful conviction that was later overturned. This highlights localized civil rights and judicial integrity concerns rather than a direct threat to national security.