Inference control in Online Analytical Processing (OLAP) systems is employed to protect sensitive data from being inferred while, at the same time, ensuring that legitimate requests can be consistently satisfied. Many models have been proposed, however most of them are suitable for only one type of aggregation, others adopted a detect-and-remove approach, which typically requires complex computations over the data and is thus too expensive to be applied in OLAP systems. In this paper, we present a practical inference control model for protecting OLAP cubes against inference attacks. PICM's has a more general framework; it is applied to any aggregation functions. In addition, PICM eliminates the source of the inference instead of detecting them. This gives a great advantage that the inference checking can, in fact, be carried out without a meaningful impact upon final query execution times.