AI Editorial Agents for Manuscript Selection in Theology Journals

Abstract

The initial manuscript selection process in scientific journals, once authors submit their proposals, is a crucial stage because it involves evaluating factors such as thematic relevance to the journal's scope, adherence to the author guidelines for the article type, and the suitability of the texts for peer review. Specifically, in the field of Theology journals, this selection is fundamental, since, although major databases like Scopus and Web of Science index them in categories such as Religious Studies and Religion, the delimitation is defined by more specific thematic criteria, such as dogmatic and doctrinal orientation, matters of pastoral or moral interest, religious practices, or beliefs. Through the selection of 41 articles published between 2023 and 2025 in the journal Cuestiones Teológicas, and the establishment of training parameters based on the journal's "Author Guidelines" and semantic fields within the area of ​​theological knowledge, an model was trained using a workflow management and automation platform (n8n) to automate the initial review process of manuscripts submitted to the journal. This article proposes the modeling of an artificial intelligence agent to verify compliance with formal requirements and the thematic relevance of manuscripts. The results demonstrate that AI-mediated journal editing, in the initial stage of the process, is not based solely on automated criteria but must be considered from the perspective of training digital selection models or editorial agents as assistants to the scientific teams responsible for selecting scholarly content.