Toss Bank Deploys AI System to Prevent Secondhand Transaction Fraud

The event involves the use of an AI system (machine learning-based predictive model) in the bank's fraud detection and prevention process. The AI system's use directly leads to harm reduction by preventing financial fraud, which is a form of harm to persons' property and financial well-being. Since the AI system's deployment has already resulted in blocking over 6,200 risky accounts and preventing approximately 3 billion KRW in losses, this constitutes an AI Incident where the AI system's use has directly led to harm prevention (a positive impact on harm). The event is not merely a potential risk or a general update but describes an active AI system in use preventing harm, thus qualifying as an AI Incident.

The article explicitly states that Toss Bank uses a machine learning-based AI model to detect and prevent fraud in financial transactions. The AI system's use directly contributes to preventing harm to customers by identifying risky accounts and alerting users before fraudulent transactions occur. Since the AI system's deployment has already prevented significant financial harm, this qualifies as an AI Incident involving the use of AI to mitigate injury or harm to persons (financial harm).

The system uses AI to analyze transaction patterns and predict potentially fraudulent accounts, issuing warnings to customers before they complete transactions. This proactive use of AI directly aims to prevent financial harm to users, which qualifies as harm to persons (financial harm). Since the AI system's use is directly linked to preventing realized or potential harm, this event involves the use of an AI system to mitigate harm. However, since the article describes the deployment and operation of the system to prevent harm rather than an incident where harm has already occurred due to AI malfunction or misuse, it is best classified as Complementary Information about AI's role in harm prevention and risk management in financial services.

The event involves the use of an AI system (machine learning model) developed and deployed to detect and prevent fraud, which is a harm to individuals' financial property. However, the article describes a preventive system designed to avoid harm rather than harm that has already occurred. Therefore, this is a case where the AI system's use could plausibly lead to preventing an AI Incident (harm), but no actual harm is reported as having occurred due to the AI system's malfunction or misuse. Hence, it qualifies as Complementary Information about a new AI system deployment aimed at harm prevention, not an AI Incident or AI Hazard.

The article describes the deployment of an AI system that uses machine learning to analyze transaction data and predict potential fraud risks in real time. This system directly aims to prevent financial harm to customers by warning them before they complete risky transactions. Since the AI system's use directly contributes to preventing harm (financial loss) to individuals, this qualifies as an AI Incident under the definition of an event where AI use has directly or indirectly led to harm (in this case, harm prevented but clearly addressed).