Living publications: How interactive artificial intelligence platforms are transforming research communication

Research communication is undergoing a paradigm shift. The traditional linear manuscript-foundational for centuries-increasingly reveals limitations in the digital era, struggling with information overload, delayed dissemination, and rigid formats. We propose a transition towards 'living publications': interactive, artificial intelligence (AI)-enhanced platforms that evolve with new evidence. Unlike static papers, these systems utilise large language models (LLMs) and vector databases to interpret context, synthesise real-time findings, and map interdisciplinary connections. This shift promises to democratise knowledge, accelerate validation, and enable dynamic evidence synthesis. However, it necessitates robust frameworks for verification, 'version of record' tracking, and peer review to maintain rigor. Successfully navigating this transition requires balancing technological innovation with preservation of academic values-ensuring that increased speed and accessibility enhance rather than diminish the quality of scientific discourse. As interactive platforms mature, they may potentially reshape how knowledge is shared, discovered, and applied, ideally accelerating scientific advancement through more dynamic, accessible, and collaborative research communication. LEVEL OF EVIDENCE: NA.