Knightscope's Park Patrol Robot Ignored Bystander Pressing Emergency Button to Alert Police about Fight

A Knightscope K5 autonomous "police" robot patrolling Huntington Park, California failed to respond to an onlooker who attempted to activate its emergency alert button when a nearby fight broke out.

Risk classification

• Primary risk domain: 5 Human-Computer Interaction
• Primary risk subdomain: 5.1 Overreliance and unsafe use

The bystander overrelied on the 'POLICE' branded robot during an emergency, expecting it to summon help, which delayed dialing 911 because the robot's emergency features were not yet active.

Additional risk subdomains

• 7.3 Lack of capability or robustness: The robot lacked the actual capability to connect to emergency services, despite its prominent branding and physical emergency button.

Causal factors

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

The risk was caused by human decisions to deploy a branded police robot with an inactive emergency button, leading to an unintentional delay in emergency response post-deployment.

EU AI Act risk tier

• Risk tier: 2 High Risk

High Risk: The system is used in law enforcement and public safety monitoring. The report describes it as a 'police robot' and 'autonomous security robot' deployed by the Huntington Park Police Department, which falls under law enforcement applications.

AI system and alleged parties

• AI system: HP RoboCop (Knightscope)
• AI purpose: Smart Surveillance; Threat Detection
• Behaviour type: Autonomous
• Alleged developer: Knightscope
• Alleged deployer: Knightscope
• Alleged harmed parties: unnamed woman injured in the fight, Cogo Guebara

Harm severity

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.

• Physical: direct Negligible, indirect Minor
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, 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 physical harm to a woman who was injured during the fight.

Directly caused: N/A

Indirectly caused: One woman suffered a head injury and bleeding from a cut during the fight, which required hospitalization; the robot's non-functional button indirectly contributed to this by delaying the emergency response by 15 minutes.

Inferred additional harm: N/A

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 5

Potential causes

Management

• Premature public deployment: Management deployed the robot to the public before activating core features.
• Prioritization of deterrence: Focus was placed on physical presence over active emergency communication.
• Misleading branding choices: Naming the robot 'POLICE' and 'RoboCop' created misleading expectations.

Technology

• Inactive emergency button: The physical alert button was not connected to the police dispatch center.
• Lack of local emergency logic: Robot continued pre-programmed patrol and music during an active emergency.
• Limited sensor integration: Cameras and intercoms were only accessible to Knightscope, not local police.

Data Inputs

• Camera feed inaccessible to cops: Live video stream went to Knightscope only, not to responding officers.
• No automated distress detection: The robot had no sensors or algorithms to detect nearby physical violence.

Human Factors

• False sense of security: Public assumed the robot was a fully functional police officer due to branding.
• Anthropomorphizing behavior: Users expected human-like response and intelligence from a simple machine.
• Overreliance on the robot: Bystanders delayed calling 911 while trying to use the inactive robot button.

Process and Methods

• Incomplete deployment protocols: Police deployed the robot on a trial basis without finishing setup.
• Lack of public signage: No signs were posted to inform the public that the button was inactive.
• Delayed relay process: Button calls went to Knightscope first instead of directly to 911 dispatch.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent, detailed accounts of the specific incident at Salt Lake Park, including statements from the bystander, the police chief, and the robot manufacturer. The facts surrounding the robot's lack of connectivity and its behavior during the incident are clear and uncontradicted.

An autonomous security robot deployed by local police failed to assist a bystander reporting a violent fight because its emergency button was inactive. While resulting in minor physical harm due to a delayed emergency response, the incident represents a minor national security concern, highlighting localized public safety risks and human-robot interaction failures rather than systemic threats.

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

Threat characteristics

• Imminence: Long-term. Represents an ongoing public safety technology integration and expectation management issue rather than an active, immediate national security crisis.
• Autonomy: Full autonomy. The security robot operated independently along its pre-programmed patrol route without human intervention or oversight during the incident.
• Novelty: Evolved capability. Demonstrates an evolved public safety challenge where deploying autonomous physical systems creates unexpected points of failure in emergency response.

Impact by dimension

• Physical security: Minor. Local public safety incident where a delayed response to a physical altercation occurred due to a non-functional robot emergency button, resulting in minor physical harm to one victim.
• Information security: Negligible. No significant intelligence compromise, information warfare, or disinformation campaigns were associated with this incident.
• Sovereignty: Negligible. The incident was limited to local municipal police operations and did not threaten state authority, territorial sovereignty, or national government functions.
• Economic security: Negligible. No strategic technology theft, financial system attacks, or threats to national economic security occurred during this event.
• Societal stability: Minor. Created a false sense of security and local public distress due to mismatched expectations of AI capabilities, but did not threaten large-scale societal stability.