Artificial intelligence in scientific research: The missing link of literacy for EU security concerns

In 2024, the EU Council Recommendation on enhancing research security, prompted by growing international tensions, stressed the importance of balancing openness in science with research security, specifically identifying artificial intelligence (AI) as a critical area. This paper argues that AI literacy is the missing link in the alignment of scientific openness with research security. In doing so, the analysis challenges the adoption of the general and misleading formula “AI in science”, proposing a distinction between AI systems as part of the research process and AI systems as research outputs. Building on this distinction, the study examines the EU legal framework on AI literacy and its application to the area of scientific research, claiming that the notion of AI literacy needs to be extended to include data literacy in science. Then, the paper frames the tension between openness and security, with AI literacy as the link (i) becoming a means of risk assessment; (ii) increasing the accuracy of research; and (iii) promoting awareness, aligned with EU values.