MotionPrint: A Client-Side Cryptographic Provenance Protocol for 3D Motion Data

The integration of Generative AI into 3D content pipelines requires verifiable provenance mechanisms. Existing watermarking techniques for 3D assets are often destructive, format-dependent, or vulnerable to removal. This paper describes MotionPrint, a cryptographic provenance protocol engineered for 3D motion interchange formats (glTF/GLB, BVH). Unlike server-centric approaches, MotionPrint implements a privacy-first, client-side architecture using WebAssembly to process assets entirely within browser memory. By implementing Ed25519 digital signatures within standard metadata containers, the protocol enables trustless verification of authorship without reliance on centralized registries or alteration of animation fidelity.