The research is aimed at studying the process of intellectualization of information systems used in the agro-industrial complex. New opportunities created by the use of artificial intelligence in agriculture are identified. The problem of the non-universality of existing testing methods and the need to develop new methods adapted for testing artificial intelligence systems is formulated. The authors have developed a model of the resulting type of testing, which takes into account the results of expert assessments, software quality characteristics, metrics and cost indicators. The authors have developed an information system on the research topic.