Inspired by mixture-of-experts models and the analysis of the hidden activations of neural networks, we introduce a novel unsupervised anomaly detection method called ARGUE. Multiple expert networks, which specialise on parts of the data deemed as normal, contribute to the overall anomaly score. For its final decision, ARGUE weights the distributed knowledge across the expert systems using a gated mixture-of-experts architecture. ARGUE achieves superior detection performance across several domains in a purely data-driven fashion and is more robust to noisy data sets than other state-of-the-art anomaly detection methods.