Recent advances in multi-material (MM) laser powder bed fusion (LPBF) have enabled the fabrication of high-performance components with spatially tailored material properties [1]. A wide range of material combinations has been fabricated and evaluated [2-5]. It has been reported that the combined effects of laser scanning strategies and powder deposition play a critical role in governing interfacial behavior and microstructural evolution in MM-LPBF systems [2]. Nevertheless, optimizing such coupled effects remains a critical challenge due to the complex interactions among multiple process parameters inherent to the MM LPBF process. Among various multi-material configurations, the Inconel 718 (IN718)/copper system represents a particularly important application. IN718 exhibits excellent mechanical strength [6], making it suitable for critical structural components, yet its relatively low thermal conductivity limits heat dissipation. In contrast, copper alloy provides superior thermal conductivity but suffers from insufficient mechanical strength. Consequently, a bimetallic IN718/Copper architecture offers a promising synergistic solution for high-temperature applications requiring both load-bearing capability and efficient heat dissipation. Despite this potential, systematic optimization of the IN718/copper interface remains underexplored, primarily due to severe thermophysical mismatches between the constituent materials. To efficiently investigate and navigate the extensive process parameter space of LPBF, machine learning (ML) has been increasingly integrated into additive manufacturing (AM) workflows [7, 8]. Among various ML paradigms, active learning employs iterative experiments–prediction loops to selectively sample high-uncertainty, informative data points, enabling efficient exploration of high-dimensional parameter spaces. Nevertheless, such strategies have not yet been systematically applied to the MM LPBF interface. In this study, an active learning framework is implemented to optimize the IN718/copper interface. By systematically examining the interplay among scanning strategies, powder distribution, and the resulting microstructural responses, this work establishes a data-driven foundation for advancing future MM-LPBF applications.