<p>Fuzzy logic is a computing paradigm that allows systems to handle imprecise or uncertain data, making them more adaptable and robust. The paper proposes the novel artificial intelligence fuzzy encryption scheme for multimedia data such as audio. With the onset of 5G and 6G development, there is a broader growth of mobile users and bandwidth-hungry wireless applications. This, in turn, mandates ultra-low latency data delivery and utmost end-to-end security as the primary concerns for wireless network designers. The constant evolution of AI and its application in cyber security is a double-edged sword, as both attackers and defenders leverage their power to outmanoeuvre each other. As computational power increases, cryptographic algorithms may need unpredictable keys to maintain security against potential attacks. So, the work proposes that the key is generated by the fuzzy transformation process, making it crucial for the decryption process. In addition, the fuzzy composition can be applied in privacy-preserving computations and secure computation over encrypted data with some degree of fuzziness, protecting individual data while still allowing meaningful results. The proposed scheme is validated against security metrics and NIST tests. It exhibits enhancements such as entropy around 14, scrambling degree equal to unity, adjacent correlation coefficient values closer to 0 at the range of 10^–5, and a large key space of 2⁹⁹⁷.</p>