Problem
Local Outlier Factor (LOF) is one of the most widely used density-based anomaly detectors: it compares the density around a point with the density around its neighbours. The cost is the neighbour search, which is roughly quadratic in brute force, while tree indices degrade in higher dimensions. On datasets with hundreds of thousands of rows the baseline becomes the bottleneck.
A second problem is interpretability: a single LOF number does not tell an analyst why a point is anomalous, or whether it is a global or a local outlier.