This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A fleet of approximately 10 Cruise autonomous vehicles became immobilized in a San Francisco intersection due to wireless connectivity issues caused by high network traffic from a nearby music festival. The incident blocked traffic and raised concerns about the potential impact on emergency vehicle access. This event occurred shortly after state regulators approved the expansion of robotaxi services in the city.
A fleet of Cruise's autonomous vehicles became unexpectedly immobilized on a busy San Francisco street, causing significant traffic disruption. The incident was attributed to wireless connectivity issues exacerbated by a nearby festival.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that fail to perform reliably or effectively under varying conditions, exposing them to errors and failures that can have significant consequences, especially in critical applications or areas that require moral reasoning.
Also classified as
The incident highlights regulatory oversight issues, as the CPUC approved unlimited fleet expansion over the objections of local transport and fire officials.
The Cruise vehicles failed to operate reliably when cellular connectivity was lost, showing a lack of robustness under real-world environmental conditions.
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The incident was caused by the Cruise autonomous vehicles' decision to immobilize themselves in the middle of the street when they lost cellular connectivity post-deployment.
A fleet of Cruise autonomous vehicles stalled in San Francisco due to cellular network congestion, highlighting a minor critical infrastructure vulnerability. While the incident caused only localized traffic disruption with no malicious intent, it underscores systemic risks associated with deploying autonomous systems reliant on public communications infrastructure.
The Cruise vehicles operate independently in real-time traffic, responding to environmental inputs. During the incident, they autonomously executed a safety fallback to stop and turn on hazard lights when connectivity was lost.
High Risk: The AI system is used in critical transport infrastructure, where failures can directly impact public safety and traffic management.
The primary function of the Cruise vehicle is autonomous driving to transport passengers without a human driver.
The report clearly details the cause of the incident (cellular network congestion from a festival), the system behavior (stalling with hazard lights), and the immediate consequences (traffic backup on a narrow street).
None
Technology
Cellular Network Dependency
Robotaxis relied on public cell networks which failed under heavy load.
No Offline Fallback Mode
Vehicles immobilized themselves completely when wireless connection was lost.
Remote Control Dependency
Safe operation and rerouting depended on remote human operator connectivity.
Data Inputs
Loss of Telemetry Data
Overloaded cell towers prevented real-time telemetry from reaching servers.
Human Factors
Crowd Network Overuse
Festival attendees overtaxed cell network, indirectly disabling AVs.
Process and Methods
Inadequate Contingency Rerouting
No effective local or offline protocol to handle vehicle connectivity loss.
Lack of Local Network Redundancy
No backup communication channels were established for high-density events.
Management
Rapid Fleet Scaling
Management expanded operations without robust communication infrastructure.
Inadequate Risk Assessment
Failed to plan for cell network congestion during large public events.
Regulatory Environment
Premature Expansion Approval
CPUC lifted restrictions despite warnings of disruptive incidents.
Absence of Fleet Size Caps
Regulators did not limit fleet size, compounding potential traffic impacts.
Lack of Mandatory Fleet Reporting
Companies are not required to report active vehicle numbers to regulators.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
