Morse code, as a dot and slash communication system, remains valuable in emergency communication due to its structural simplicity. However, traditional Morse code usually uses sound signals, which can only propagate short distances and are easily affected by noise. Here, we present an intelligent Morse code classification system integrating flexible iontronic pressure sensor with artificial intelligence (AI) analysis. The iontronic pressure-sensitive mechanism employs the supercapacitor interface based on the electric double layer (EDL) effect. To improve sensor sensitivity, the sensitive layer is designed with a conical microstructure, and deformation process is validated through finite element analysis. Experimental results demonstrate a maximum sensitivity of 1821.1 kPa−1 within the 0–300 kPa range, coupled with fast response times of 11 ms and 8 ms. A high-speed capacitive readout circuit and wireless transmission enable system miniaturization and real-time operation. Morse code dot and dash signals are mapped through the impact and static pressure responses of sensor, respectively, forming a CNN classification model. The CNN model achieves 93.6% accuracy in classifying Morse code digits 0–9, validated by t-SNE and confusion matrix visualization. This intelligent system has significant potential in emergency communication during hazardous situations.