Gunshot Detection Technology ShotSpotter (now SoundThinking) Reportedly Only Has 47% Accuracy in Chicago System

SoundThinking's (formerly ShotSpotter's) system in Chicago, with a reported 47% accuracy rate for detecting actual gunshots, led to potential public safety risks by failing to alert police to real shootings, possibly delaying emergency response to violent incidents and misdirecting law enforcement resources.

Risk classification

• Primary risk domain: 7 AI system safety, failures, & limitations
• Primary risk subdomain: 7.3 Lack of capability or robustness

The ShotSpotter system failed to perform reliably under operational conditions, achieving only a 47% accuracy rate and missing 72 actual shooting incidents.

Additional risk subdomains

• 1.1 Unfair discrimination and misrepresentation: The technology was disproportionately deployed in Black and brown communities, leading to concerns about systemic bias and dangerous over-policing in these neighborhoods.
• 5.1 Overreliance and unsafe use: The Chicago Police Department continued to rely on and expand the system despite internal data showing it missed more than half of the shootings where people were struck.

Causal factors

• Entity: AI
• Intent: Unintentional
• Timing: Post-deployment

The incident was caused by the ShotSpotter AI system failing to perform its intended function of detecting gunshots, which was an unintentional and unexpected failure post-deployment.

EU AI Act risk tier

• Risk tier: 2 High Risk

High Risk: The system is used in law enforcement applications, specifically for gunshot detection and police deployment, which is explicitly classified as high risk under the EU AI Act.

AI system and alleged parties

• AI system: ShotSpotter (SoundThinking)
• AI purpose: Gunshot Detection; Threat Detection
• Behaviour type: Tool
• Alleged developer: SoundThinking, ShotSpotter
• Alleged deployer: Chicago Police Department
• Alleged harmed parties: General public, Chicago Residents, Chicago Minority Communities

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Severe. 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.

• Physical: direct Negligible, indirect Substantial
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Substantial, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Substantial, indirect Substantial
• Civil rights: direct Negligible, indirect Minor
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Severe, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Physical

Reported: The report explicitly describes instances where people were struck by gunfire but the system failed to detect the shots, potentially delaying emergency response.

Indirectly caused: The system failed to detect 72 instances where a person was struck by gunfire, leading to delayed emergency medical response for those victims.

Inferred additional harm: It is highly likely that the delayed emergency response for some of the 72 victims led to worsened physical injuries or preventable fatalities, though specific numbers are not detailed.

Financial

Reported: The report describes a three-year contract extension worth $33 million signed in 2018, a $10 million payment in 2023, and a $9 million budget allocation for 2024.

Directly caused: The city of Chicago spent $33 million on a contract for a system that performed significantly below its promised accuracy rate of 80-90% in the tracked district.

Inferred additional harm: The continued allocation of millions of dollars to an ineffective system represents a significant opportunity cost, diverting public funds away from community-led safety solutions, housing, and healthcare.

Differential treatment

Reported: The report explicitly describes the disproportionate deployment of the technology in Black and brown communities.

Directly caused: The gunshot detection sensors were specifically deployed in long-divested and heavily segregated South and West Side neighborhoods.

Indirectly caused: This targeted deployment subjected minority residents to higher levels of surveillance and potentially dangerous police responses compared to other neighborhoods.

Inferred additional harm: The systemic concentration of error-prone surveillance tools in minority neighborhoods likely exacerbated existing racial disparities in policing and law enforcement outcomes.

Civil rights

Reported: The report explicitly mentions concerns from community groups that the system results in deployments that are dangerous for Black and brown community members.

Indirectly caused: The deployment of police officers responding to unverified or false acoustic alerts increases the risk of hostile police encounters for minority residents.

Inferred additional harm: The continuous acoustic monitoring of public spaces violates the general expectation of privacy and civil liberties for millions of residents in targeted neighborhoods.

Privacy

Reported: The report explicitly describes a vast network of microphones mounted on lampposts and utility poles to monitor public spaces.

Directly caused: The system continuously monitors and records acoustic data across a 100-square-mile footprint in Chicago, capturing environmental sounds without resident consent.

Inferred additional harm: The widespread deployment of acoustic sensors raises significant privacy concerns regarding the potential interception and recording of private conversations in public spaces.

People affected

• Occurrences reported: 1
• People reportedly harmed: 72
• People reportedly exposed: 135

Potential causes

Management

• Ignoring Performance Reports: Leadership signed contract extensions despite known 47 percent accuracy rate.
• Prioritizing Tech Rebranding: CPD prioritized high-tech policing to rebuild trust amidst legitimacy crisis.

Technology

• Low Gunshot Detection Accuracy: System only detected 47 percent of shootings where a person was struck.
• Outdoor Gunfire Design Limitation: System is designed only for outdoor gunfire, missing indoor shootings.

Data Inputs

• Flawed Validation Data: Accuracy studies relied on self-reported police data instead of field tests.
• Acoustic Signal Classification: Validation focused on sound classification rather than overall detection.

Human Factors

• Leadership Overreliance: Superintendents defended technology despite documented performance failures.
• Lack of Operator Understanding: Officers and public misunderstood how acoustic detection actually works.

Process and Methods

• Inadequate Performance Tracking: CPD lacked standardized processes to track and map missed gunshots.
• Flawed Study Methodology: Commissioned studies failed to conduct independent field testing.

Regulatory Environment

• Lack of Independent Oversight: No regulatory body audited the performance claims of the surveillance system.

Information quality

• Classification confidence: High
• Reason for confidence: The report provides highly detailed, empirical data from internal Chicago Police Department tracking spreadsheets, including specific accuracy rates (47%) and the exact number of missed shooting incidents (72 out of 135). It also details the financial contracts and community responses, leaving very little ambiguity about the system's performance and deployment context.
• Ambiguities identified: The report does not explicitly state the medical outcomes of the 72 victims whose shootings were missed, leaving the exact physical harm (fatalities vs. injuries) to be inferred.

The performance failure of the ShotSpotter gunshot detection system in Chicago represents a local public safety and civil liberties concern rather than a national security threat. While the 47 percent accuracy rate and disproportionate deployment in minority neighborhoods raise significant municipal policy, privacy, and human rights questions, the incident does not impact national critical infrastructure, sovereignty, or strategic security.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents a localized, long-term public safety and technology efficacy concern rather than an imminent national security threat.
• Autonomy: Human-controlled. The acoustic detection system alerts human operators and police dispatchers who make all tactical deployment and emergency response decisions.
• Novelty: Established threat. Acoustic monitoring and localized sensor failures represent established technology types with known deployment challenges and limitations.

Impact by dimension

• Physical security: Negligible. Local public safety failure regarding gunshot detection in Chicago does not threaten national critical infrastructure or constitute a national kinetic security threat.
• Information security: Negligible. The incident involves local law enforcement acoustic sensors and does not compromise national intelligence capabilities or involve information warfare.
• Sovereignty: Negligible. Local municipal procurement and police operations do not impact national sovereignty or federal government decision-making processes.
• Economic security: Negligible. The financial impact is limited to a local municipal contract of 33 million dollars and does not threaten national economic stability or strategic technology assets.
• Societal stability: Minor. Disproportionate deployment in minority neighborhoods raised civil rights and surveillance concerns, but impact remains localized to municipal policing rather than national-scale instability.