This study reconsiders responsible artificial intelligence (RAI) from a perspective of business. Machine learning-based artificial intelligence systems have widely been utilised for various purposes due to extremely useful functions such autonomous systems can perform. On the other hand, however, the autonomy tends to make system behaviour unpredictable and uncontrollable, and thus to make it difficult to determine exact cause as well as to define the locus of responsibility when the systems bring any harm. In this respect, RAI becomes an ethical and social matter. Although a variety of principles and recommendations for RAI have been offered, they lack a perspective of business which is critical to ensuring the effectiveness of them. Efforts to compensate for the lack are urgently necessary to develop sound economy and society which have increasingly been dependant on AI technologies.