In recent years, High-Altitude Wind Energy (HAWE) technologies have emerged to harness the power of wind blowing up to 1000 m above the ground, exploiting the controlled flight of tethered wings. Automatic control of the wings is a key aspect of HAWE generators and it is a complex issue, since the system at hand is open-loop unstable, difficult to model and subject to significant external disturbances. In this paper, a control technique for HAWE generators is proposed, based on a Direct-Inverse Control approach, that allows to compute a controller directly from data, thus avoiding the need to derive a model of the system. The Direct-Inverse controller is designed by using a sparse identification method, and its effectiveness in the considered application is shown through numerical simulation results.