Ubisoft uses machine learning tools built by its La Forge R&D group to speed up the conversion of raw motion capture recordings into finished character animations, reducing a task that can take a human animator approximately four hours down to roughly four minutes. A separate tool called ZooBuilder generates animal animation skeletons from video footage, removing the need for motion capture sessions with live animals.

Raw motion capture (MOCAP) data typically requires significant manual cleanup before it looks natural in a game. Ubisoft La Forge published peer-reviewed research at SIGGRAPH 2020 introducing a system called Robust Motion In-Betweening, which uses machine learning to automatically fill in missing animation frames between keyframes. ZooBuilder, originally developed at Ubisoft's China AI and Data Labs and presented at the Symposium on Computer Animation (SCA) 2020, takes 2D video footage of animals as input and outputs 3D animated skeletons; Ubisoft has noted that for species like mountain lions and elephants, traditional manual keyframe animation of a single animal could require up to 500 person-days. As of a 2022 public demonstration, ZooBuilder was described as a prototype not yet deployed in shipped productions. Separately, La Forge has also developed tools for automatically generating realistic human face textures and lip-sync animations for different language dubs. These tools are used internally across Ubisoft productions; specific shipped titles using each tool have not been individually confirmed.