Open AI services for making library information analytical products

The author discusses how the AI can be utilized to prepare information analytical products (IAP) in scientific libraries. Traditional methods of IAP preparation have become too labor-consuming in the context of science information burst. Russian libraries are challenged by financial problems when implementing expensive proprietary AI-based software, as well as by sanction limitations affecting many online resources of advanced functionality. The purpose of the study is to identify and classify the functionality of AI services available for Russian users to optimize IAP acquisition at science libraries. The author analyzed 80 AI services and selected those with open and conditionally open access. Based on the technological stages of information analytico-synthetical processing, the author suggests the typology for several groups of services: 1) for information retrieval (Perplexity AI, Semantic Scholar, Google Alerts); 2) for data analysis, synthesis, and visualization (ResearchRabbit, Elicit, Napkin AI); 3) generative tools for acquiring additional material (texts, images, audio, and presentations).. The revealed AI services may be implemented in science libraries to improve quality and accelerate the preparation processes for reviews, digests, dashboards, and other IAPs. The author specifies that the AI technology performance depends directly on librarian’s competences, his or her ability to assess critically retrieved responses, while AI implementation carries the risks of data confidentiality breach, hallucinations, etc. The originality of the study lies in the analysis of the AI open segment available for Russian libraries under current conditions, and the typology developed for the IAP preparation full cycle.