The increase in technologies at the customer interface is almost necessary to increase the efficiency of service production. Alongside customer service additional support has been sought from different artificial intelligence technologies, especially chatbots, but their use and the factors affecting it have still been studied only briefly in the literature. The aim of this study was to map the factors influencing the adoption of chatbots, as well as their strengths and weaknesses as an interaction channel. The study utilized an exploratory perspective, in addition to which the theory of adoption of artificial intelligence technologies was used as a theoretical frame of reference. The explanatory variables in that are clarity of roles, ability, motivation, individual differences, innovation characteristics and experiences of trust, privacy and suspicion. The study was conducted as a structured online survey with totality of 426 responses. In addition to quantitative data, the study collected qualitative data through open-ended questions. The significance of the measured variables for adoption was examined through multinomial and binary logistic regression analysis. Two regression models were constructed in the study, one describing the factors influencing the adoption of chatbots and other for factors influencing the choice of preferred interaction channel. Qualitative material was analyzed with thematising and mind maps. As a result of the empirical study, it was concluded that the theoretical framework utilized partly explained the adoption of chatbots. Based on the regression model, factors positively influencing acceptance were external motivation, internal motivation, relative advantage, suitability and ease of use. However, shortcomings were also found in theory. Based on the analysis of qualitative data, the adoption was negatively affected by the lack of individuality, uncertainty and distrust of chatbots, lack of humanity, lack of technological development and the challenge of assessing the ability of chatbots from the customer's point of view. According to the regression model, the choice of transaction channel was influenced by internal motivation, suitability and the need for interaction. The strengths and weaknesses of chatbots identified on the basis of qualitative data partially supported the previous research. However, new observations were also made about the weaknesses, which show the value of technology and the importance of communicating its role from the customer's point of view.