Skeletal animation is the standard technique for animating 3D characters and creatures, using a hierarchical system of interconnected bones to drive the deformation of a surface mesh. Rather than directly manipulating every vertex of a character's geometry, animators pose and move the underlying skeleton, and the mesh deforms automatically to follow the bone positions, creating natural organic movement from a manageable number of controls.

The skeleton consists of a hierarchy of joints connected in parent-child relationships that mirror the structure of the animated subject - a human character's skeleton might include spine, shoulder, arm, hand, hip, leg, and foot joints in a chain where moving a parent joint carries all its children along with it. Each vertex of the character mesh is assigned influence weights to one or more bones, determining how much each bone's movement deforms that area of the surface. Skinning or weight painting is the process of assigning and refining these influence values to produce natural deformation at joints. Skeletal animation is used across game engines, film production, and interactive media because of its computational efficiency, editability, and compatibility with motion capture data, which can be applied directly to a compatible skeleton to drive character animation from recorded performance.

Understanding skeletal animation provides context for the difference between AI-generated character motion and traditionally produced character animation. In AI video generation, motion is synthesized probabilistically from training data patterns rather than calculated from a skeleton, which enables rapid generation but results in different characteristics and limitations - particularly around precise physical interactions and extreme poses where skeletal systems with inverse kinematics would produce more reliable results.