Cloud computing has revolutionized modern computing by providing scalable and on-demand computing resources. However, efficient resource allocation remains a critical challenge, directly affecting system performance, cost, and energy consumption. This research explores the role of AI-driven predictive load balancing in optimizing resource allocation within the scope of performance engineering and cloud engineering. By leveraging machine learning-based forecasting models, cloud infrastructure can predict workload fluctuations and dynamically allocate resources, improving efficiency, reliability, and overall system performance. Experimental results demonstrate significant improvements in resource utilization, response time, and energy efficiency