Wind power forecasting is essential for the integration of wind power producers (WPPs) into existing energy grids and energy markets. For decades, numerical weather prediction (NWP) has formed the cornerstone of wind power forecasting efforts. Due to the high computational requirements and technical complexities associated with NWP, WPPs typically purchase their forecasts from external service providers. However, with the arrival of advanced statistical methods and artificial intelligence weather prediction (AIWP) techniques, this landscape is rapidly transforming. These modern forecasting approaches are encroaching on traditional NWP in terms of accuracy and anomaly prediction for medium-range forecast horizons, which may usher in a new era of wind power forecasting. This review article explores the characteristics of a valuable wind power forecast and investigates current NWP models to reveal limitations that could be solved with AIWP. As part of this investigation, AIWP models are compared to NWP to identify the key models responsible for this revolution. It reveals that AIWP is improving rapidly compared to NWP, suggesting a potential integration of AIWP as a decision-making tool. While AIWP models do not yet surpass the accuracy of traditional physical models across all validation criteria, they provide cost-effective and computationally efficient alternatives that are less complex to implement. The advantages and concerns of AIWP highlight the readiness of these AIWP models to be implemented in the wind power industry.