Iris recognition systems are crucial for secure biometric verification; however, they are vulnerable to spoofing attacks using false representations such as contact lenses or printed irises. Discrete Wavelet Transform (DWT) and Discrete Cosine Transform (DCT), two powerful feature extraction algorithms, are used in this study to create a novel hybrid approach for iris liveness detection. These techniques improve the model’s capacity to discriminate between real and artificial irises by obtaining data in both the frequency and spatial domains. A group of deep learning models, such as ResNet and EfficientNet, are used to classify actual and artificial irises. While EfficientNet’s scalable architecture maximizes efficiency and minimizes computational complexity, ResNet’s residual connections enable accurate deep feature learning. The proposed approach achieved a remarkable accuracy rate of 100% in detecting printed spoof data and 99.18% with patterned spoof data on a comprehensive dataset, underscoring its effectiveness in bolstering biometric security against various spoofing techniques. This hybrid DCT-DWT feature extraction combined with the EfficientNet-ResNet ensemble significantly improves the reliability of iris identification systems, positioning it as a viable solution for real-world biometric authentication challenges. Future research could extend these methods to accommodate additional spoofing patterns and adapt the model for other biometric modalities.