Full-Scale Operation of AI-Powered Benefit Fraud Detection Begins

Neo First Life Insurance Co., Ltd. Neo First Life Insurance Co., Ltd. (President: Takashi Uehara; hereinafter "the Company"), a member of the Dai-ichi Life Group, announces that it has completed the development and implementation of its "AI-supported initiative to enhance benefit fraud detection," as previously disclosed in the news release dated June 16, 2025. Starting in April 2026, these mechanisms will be integrated into actual business processes for full-scale operation. 【Overview of the Initiative】 The system introduced in October 2025 uses AI to organize and analyze trends and commonalities in past benefit claim data. By automatically extracting "claims with unusual characteristics" or "patterns similar to past cases," the system enables a broader range of detection, including signs that were previously difficult for human eyes to notice. Furthermore, the system finalized in December 2025 employs an ensemble learning approach, which combines multiple AI models to increase accuracy. Specifically, it utilizes a two-stage configuration: ・A model group that prioritizes broad detection to ensure even slight signs of suspicion are not overlooked. ・A model group that identifies claims requiring closer attention and supports the prioritization of responses. This dual approach achieves both "widening the detection net" and "improving the accuracy of judgment." In the full-scale operation beginning in April 2026, these efforts will reduce the time from AI detection to the assessment of response necessity by expert staff, while also accelerating the entire process following those assessments. 【Overall Structure of the Initiative】 【Benefits of the Initiative】 This initiative is designed to have AI support the decision-making of expert staff. By utilizing the analysis results and priority levels provided by the AI, staff can quickly identify claims that require attention, enabling more precise and rapid decision-making. Expected benefits include: ・Rapid payment for legitimate claims. ・Early response to potentially fraudulent claims. ・Reduced lead time for determining response policies after assessment. The company has already achieved AI-driven detection and priority assessment for certain moral risks. 【Future Outlook】 The Company is advancing this initiative to maintain the insurance system in a sound and fair manner. Benefit claims are an essential system for ensuring that customers who truly need coverage can use it with peace of mind, and the Company prioritizes protecting the reliability of this system above all else. Furthermore, the Company will continue to deepen its efforts by combining human expertise with AI, aiming to improve quality and further enhance the customer experience (CX).