Abstract: After gravity-induced deformation is coupled with manufacturing deviations, the geometric accuracy of composite fuselage panels often fails to meet assembly requirements; therefore, shape control is essential in aircraft assembly. Given the high brittleness and vulnerability of composite panels under loading, assembly force levels must be constrained while correcting shape. However, the complex thin-walled structure and uncertainties in the assembly process make the deformation behavior and load state difficult to characterize, complicating shape regulation. This paper proposes an uncertainty aware force–shape synchronous control method. The method employs a Gaussian process that explicitly accounts for uncertainties as the predictive core, and couples it with multi-objective optimization (NSGA-II); a reference point selection scheme is used to output executable optimal displacements of the actuators. To evaluate effectiveness, validation is conducted on a 3 m × 2 m composite fuselage panel. The results show that, compared with a Gaussian process that does not consider uncertainties, the proposed model yields a significantly lower average relative prediction error for both deformation and assembly force; after shape regulation, the maximum deviation is reduced from 2.05 mm to 0.29 mm, while the per-axis assembly force/torque are controlled within 40 N and 50 N·m, respectively. These results demonstrate that the method effectively reduces shape deviations and maintains low assembly forces of composite fuselage panels, providing a reusable engineering framework for low-damage, high-precision assembly of thin-walled composite components.