Design of white circularly polarized luminescence with high glum across entire visible regions by dual-loop active learning

White circularly polarized luminescence (WCPL) materials have attracted increasing attention. However, achieving WCPL simultaneously with high color rendering index (CRI) value at ideal CIE coordinates and large dissymmetry factor (glum) across entire visible regions remains a significant challenging, due to the vast chemical design space and the inefficient experimental screening by human-driven intuition-based methods. Herein, we propose an AI-assisted interactive experiment–learning evolution strategy to accelerate the discovery of WCPL materials with optimal trade-offs, integrating model recommendation, experimental validations, and active learning in each iterative refinement cycle. A multi-objective optimization algorithm automatically guides the experimental design with a dual-loop active learning framework, one for maximizing CRI at ideal CIE coordinates, and the other for optimizing glum value across entire visible regions. High-performance WCPL materials with excellent |glum | (>1.5) range of 410 to 680 nm, high CRI ( > 90) at ideal CIE coordinates (0.33, 0.33) can simultaneously be achieved with limited experimental date. Moreover, this workflow also enables on-demand fabrication of high performance white circularly polarized organic light-emitting diodes as well as the user-specified correlated color temperature, facilitating their application in lighting and full-color 3D display. White circularly polarized luminescence materials are promising but combining favorable white-light colorimetric properties and strong chiroptical activities is difficult. Yang et al. adopted a dual-loop active learning strategy to realize high-performance materials and devices.