The article deals with the problem of using artificial intelligence (AI) in the construction sector, since these systems are not being actively implemented, though AI capabilities can significantly simplify the entire process of erecting buildings and structures, including the development of improved business plans, adjusting the progress of the project directly on the construction site, making suggestions for management decisions in case of deviations from schedules and logistics plans. The experience of Chinese and domestic construction organizations in this area has been studied. A scheme for the implementation of an artificial intelligence system in construction production has been developed, the necessary resources for the development of machine learning in this area have been analyzed. It is concluded that for correct decision-making it is necessary to apply the principles of discrete mathematics, in particular, the distribution of expert assessments accepted as linguistic variables on the Harrington desirability scale with the construction of the membership function.