In this study, the damage performance of carbon fiber reinforced polymer (CFRP) strengthened Q345 steel tubes under bending loads was researched. Through three-point bending tests, the bending strength and energy absorption performance under different reinforcement methods were evaluated using energy characteristic analysis. Additionally, using acoustic emission (AE) techniques, the reinforcement effects of different CFRP layup methods on steel tubes were comparatively analyzed, as well as exploring the evolution of acoustic characteristics of internal damage and bending failure. Finally, a damage classification model optimized by the bat algorithm (BA) for the least squares support vector machine (LSSVM) was proposed. The study finds that CFRP winding layers increasing can significantly enhance the bending strength and energy absorption capacity of the steel tubes, but increasing the winding angle will reduce the structural performance. By comparing the acoustic emission signals of specimens under different reinforcement methods, the effectiveness of acoustic emission technology in revealing the damage modes of carbon fiber reinforced steel tubes during bending was confirmed. Analysis of energy probability density and maximum likelihood estimation shows that, the composite tubes acoustic emission energy follow a power-law distribution at different energy levels, with the energy distribution exponent increasing with the increase of CFRP winding layers and decreasing with the increase of winding angle. The BA-LSSVM model was established to classify the degree of damage during the specimens damage process, with an accuracy of over 98%.