What are the changes that the development of AI brings to the character of labor status in organizations? Are there new momenta for cooperation between staff in organizations? Does AI initiate transformations in the division of labor? The paper addresses these questions on the basis of field research organized and conducted in Russia and Belarus. Employees of 16 Russian and Belorussian IT companies were interviewed. The foci of the paper are the new structures of labor occupation and the new types of expertise that have appeared on the scene with AI entering into the everyday life of society. The author discusses three basic trends. First, the appearance of new categories of activities and new types of expertise in the frame of cooperation between employees. Second, the emergence of "invisible" varieties of labor. Third, the replacement of local knowledge with the knowledge that stems from 'big data'. The main result of the study is the following. The division of labor in organizations influenced by the development of artificial intelligence / machine learning is characterized by a complex and specific type of cooperation which has a structure comparable to a "Russian doll". In order to portray this structure the author applies the concept of interactional expertise proposed by Harry Collins. The paper also formulates several research problems and hypotheses connected with current transformations in the division of labor, as well as with new experiences of employees who interact with AI technologies.