TY - JOUR TI - Developing Effective and Value-Aligned AI Tools for Journalists : 12 Critical Questions to Reflect upon AU - Cordero, José Miguel AU - Henn, Theresa AU - Holtel, Frederik AU - Sánchez Gómez, José Á. AU - Arenas, Diego AU - Šipka, Andrea AU - Vollmer, Sebastian PY - 2026 JO - Journalism Practice VL - 20 IS - 9 SP - 2776-2795 PB - Routledge, Taylor & Francis CY - London SN - 1751-2794 DO - 10.1080/17512786.2025.2465894 UR - https://fis.uni-bamberg.de/handle/uniba/107450 LA - eng AB - Artificial intelligence (AI) is playing an increasingly prominent role in journalism, with emerging tools like large language models (LLMs) and other AI applications gaining traction. This paper explores the current landscape in adopting AI tools for the information gathering and story development phase in journalistic practice. We identify two contemporary challenges for implementing AI tools in journalism: effectiveness and value alignment. While the first challenge includes the issues of generalizability, costs, user-friendliness, and accuracy, the latter refers to adherence to journalistic values, independence from external actors, transparency, and data privacy issues. We reflect upon these challenges based on the development process of prototyping an open-access AI tool for classifying documents and prioritizing information-rich pages in collaboration with the Organized Crime and Corruption Reporting Project (OCCRP), a globally recognized investigative journalism network. Based on the literature and our experience, we derive twelve critical questions that we deem important to guide the development of future AI tools for journalism. KW - Journalistic AI KW - generative AI KW - investigative journalism KW - computational journalism KW - document classification KW - machine learning ER -