This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Waymo's autonomous vehicles were involved in multiple incidents where they failed to stop for school buses with active safety signals, including instances where students were actively crossing the road. Despite software updates intended to resolve the issue, violations continued, leading to a formal NHTSA investigation and a voluntary software recall by Waymo. No injuries were reported, but regulators and school districts expressed significant concern regarding the ongoing safety risk to students.
Officials in Texas reportedly claimed that Waymo's self-driving vehicles illegally passed stopped school buses at least 19 times during the 2025 school year. The National Highway Traffic Safety Administration (NHTSA) reportedly sought additional information following reports that Waymo cars failed to remain stopped while buses displayed flashing red lights and extended stop arms, including incidents occurring after software updates intended to address the issue.
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.
The incident represents a lack of capability and robustness in the Waymo autonomous driving system, which failed to perform reliably under standard traffic conditions involving stopped school buses.
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 safety risk was caused by a technical failure in the Waymo autonomous driving software, which unexpectedly failed to stop for school buses after being deployed on public roads.
Waymo self-driving vehicles repeatedly failed to stop for school buses with active safety signals in Texas and Georgia, leading to a NHTSA investigation and a voluntary software recall. While presenting a localized public safety concern for school children, the incident is an unintentional technical failure with negligible national security implications across all assessed categories.
The Waymo vehicles operate independently on public roads, making real-time driving decisions such as slowing, stopping, or passing school buses based on environmental inputs without human intervention.
High Risk: The AI system is used in critical transport infrastructure and safety-critical automotive applications, directly impacting public safety.
The AI system's primary goal is to pilot self-driving vehicles safely and autonomously on public roads.
The reports provide consistent and detailed accounts of the incidents from multiple sources, including school district officials, police departments, Waymo representatives, and the NHTSA. The specific number of violations, locations, and the nature of the software failure are clearly documented.
None. The incident is clearly a technical failure of an autonomous vehicle's perception and decision-making system.
Technology
Defective Stopping Logic
Software caused vehicles to stop and then resume driving near school buses.
Failed Software Update
The November software patch did not stop vehicles from passing school buses.
Data Inputs
Side Street Detection Limits
Approaching from side streets hindered proper detection of bus safety signals.
Process and Methods
Inadequate Update Validation
Software updates were released without verifying they resolved the safety risk.
Real-Time System Development
Complex systems were developed and tested in real time on public roads.
Management
Refusal to Pause Operations
Management rejected school requests to halt service during pickup times.
Delayed Recall Decision
Recall was only filed after multiple violations and federal investigations.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
