Yiren Digital Ltd. (NYSE: YRD) announced progress in building a shared enterprise AI operating model that allows AI capabilities developed in one business function to be deployed across others without rebuilding core models or workflows, according to a press release issued Aug. 18, 2026.

The Beijing-based financial technology company said its common enterprise AI framework standardizes models, agents, workflows and governance across functions including customer operations, marketing, capital operations, risk management, compliance and research and development.

Key components of the architecture include two proprietary large language models called Zhiyu and Yizhi, a multi-agent platform called MagiCube 2.0, a workflow execution engine called XuanJi and an orchestration layer called ZhiNao.

The company said its credit and insurance businesses have served as initial production environments for developing and validating these AI capabilities before broader deployment across the organization.

"Our architecture prioritizes module and model reusability – every fraud detection model, validation framework, and decision layer is built as a composable component," said Ning Tang, Chairman and Chief Executive Officer of Yiren Digital. "As the AI agent ecosystem matures, our fraud detection capabilities can be seamlessly integrated into autonomous workflows. Ultimately, we are building the technology for autonomous financial system."

The company said the shared framework is intended to shorten development cycles and reduce implementation time for future AI deployments. Yiren Digital added that it plans to extend standardized AI capabilities into additional application-layer opportunities beyond its existing businesses.