safety-critical AI
AI whose malfunction could cause serious harm to people or property.
Definition
Systems whose failure or malfunction may lead to significant bodily injury, death, or substantial property damage in road-safety contexts (e.g., traffic-control automation or automated vehicle interventions). The bill treats these systems as high-risk and subjects them to stricter testing, certification, monitoring and human-oversight requirements.
Related Terms
Safety-impacting AI
AI applications that could create or exacerbate physical or safety risks....
High‑Risk Artificial Intelligence
AI systems materially affecting life, safety, or fundamental rights....
Autonomous AI System (SIAA)
AI systems capable of operating without direct human intervention....
High-risk AI systems
AI systems subject to elevated regulatory controls....
High-risk AI applications
AI systems that significantly affect rights, safety, or essential services....