Artificial intelligence (AI) is poised to become the key technology of the future. This technology enables a machine to simulate human behaviour, such as reasoning, planning and creativity by learning algorithms. Companies are increasingly turning to AI to develop automated, personalized, “client-centric” services. AI also allows for new opportunities to strengthen and facilitate risk and fraud detection and reduction as well as to promote regulatory compliance. However, using it generates risks as illustrated by Deloitte in one of its research studies (1): › Quality, quantity and relevance of data used. The results of AI systems depend on the quality and quantity of data. If the dataset used to build algorithms contain biases, the algorithm generated will likely also reflect those biases, and even amplify them. › Opacity (black box effect in the context of AI). In contrast to previous generations of AI where systems would make very transparent decisions that were established by humans, the new generations rely on very complex statistical methods, based on thousands of parameters. All these factors will make final decisions difficult to interpret, or even impossible to explain by humans. › Potential for dysfunction. Algorithms do not have the conceptual understanding capacities and common sense that humans possess, and which are necessary to assess radically new situations. As recent reports by the World Economic Forum have underscored, artificial intelligence is at a turning point. Short term, it seems important for the development of artificial intelligence to satisfy minimum governance, ethics and risk management requirements. According to Deloitte (2), this discussion should focus on proving the reliability of the algorithms used (from the standpoint of their internal and external auditing), the intelligibility of models, and interactions between humans and intelligent algorithms. It is therefore crucial that the corporate governance committee’s mandate be revised to include this discussion and develop a code for using artificial intelligence that will assure shareholders and interested parties (stakeholders) that it is being developed and used by putting humans at the centre of the machine, by guaranteeing the accuracy, security and confidentiality of the data fed into it, and by regulating algorithms to integrate diversity and override decision-making biases.