Crisis-response capacity is the ability of an AI system or platform to recognize signs of serious user distress or imminent harm and respond appropriately. It may include escalation pathways, human support, safety messaging, and safeguards against harmful interactions.