Apple's crash detection feature in iPhone 14 and Apple Watch Series 8 devices has been generating thousands of false 911 emergency calls from skiers and snowboarders, overwhelming emergency dispatch centers and diverting resources from real emergencies.
Apple launched the iPhone 14 and Apple Watch Series 8 in September 2022 with crash detection software designed to automatically call 911 when users are in vehicle accidents. The feature gives users 20 seconds to cancel before placing the emergency call. However, the system has proven overly sensitive to skiing and snowboarding activities. Summit County, Colorado reported 185 false calls in one week from January 13-22, with emergency dispatchers receiving up to 20 false calls per day. Greene County, New York saw emergency calls increase 16% in December 2022 compared to 2021. Parks Canada received approximately 30 false calls from Jasper National Park's Marmot Basin ski area as of January 24. The false alarms have overwhelmed 911 dispatch centers, forcing operators to put real emergencies on hold while investigating automated calls. Each call requires verification and may trigger ski patrol responses. Apple released iOS 16.2 updates in December 2022 to reduce false detections by requiring proximity to roads, and the company sent representatives to observe dispatch operations. While the feature has successfully helped in real emergencies, including saving an unconscious Indianapolis driver, the false positives have strained emergency response systems across North American ski regions.
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.
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