A new base for mathematical modeling in the field of information ethics is proposed. A concept of the so-called biased derivative is introduced. It has a potential to investigate dynamical properties of information processing systems where rules of ethics are formulated.