This study explores AI-based encryption techniques for securing data transmission in telecommunication systems, addressing the growing need for robust cybersecurity measures in an era of increasing cyber threats and data breaches. Traditional encryption methods, while effective, often suffer from computational inefficiencies, vulnerability to evolving attacks, and challenges in key management. By leveraging artificial intelligence, particularly machine learning and deep learning algorithms, this research presents an adaptive encryption framework capable of dynamically enhancing security measures while optimizing computational performance. The proposed AI-driven encryption model integrates predictive analytics for threat detection, automated key generation, and intelligent encryption mechanisms to improve data protection against unauthorized access and cyberattacks. Experimental results demonstrate significant improvements in encryption speed, data integrity, and resilience against various cryptographic attacks, while also reducing computational overhead and energy consumption. The study further highlights the adaptability of AI-driven encryption in responding to emerging cybersecurity challenges, ensuring secure, real-time communication in telecommunication networks. The findings underscore the potential of AI in revolutionizing cryptographic approaches, offering a scalable, efficient, and intelligent security framework for modern telecommunication infrastructures. Future research should focus on refining AI-based encryption techniques by integrating blockchain, federated learning, and hybrid cryptographic models to further enhance security, privacy, and efficiency in data transmission.