<p>Soft robotics offers a venue to narrow the gap in manoeuvrability and efficiency between engineered vehicles and swimming or flying animals. Yet, state estimation and control of highly deformable structures remain challenging, leaving soft robots vulnerable to unsteady environmental flow disturbances. Inspired by animals’ ability to sense and respond to fluid forces via appendage shape changes, we demonstrate a soft robotic wing with a flexible proprioceptive e-skin that autonomously detects and compensates for sudden disturbances. Experiments show that while the wing’s passive elastic compliance alone mitigates lift deviation compared to a rigid wing, it still leaves a large unwanted lift bias. By integrating proprioception and active shape morphing, we establish a hybrid passive-active disturbance rejection strategy in which passive material compliance reduces baseline deviations and active control suppresses residual biases. This combination autonomously reduces the unwanted lift impulse over the disturbance by 87%, closely matching the gust-rejection abilities of some flying animals. These results demonstrate how embodied intelligence and hybrid control could naturally endow soft robots with disturbance-resilient capabilities akin to those of living organisms.</p>