The science publishing manifesto: AI moves fast, science publishing must too

Purpose We are witnessing rapid and unprecedented developments in artificial intelligence (AI), exposing the vulnerabilities of the current scientific publishing ecosystem. This system – comprising, among others, commercial and non-profit publishers, scholarly journals, editors and reviewers – struggles to keep pace with the speed of AI research. The paper calls for urgent reform among the key actors shaping how manuscripts are handled: publishers, journal editors, editorial boards and reviewers. Design/methodology/approach The article examines the ecosystem’s role in shaping research dissemination. It contrasts the pace of the AI industry – especially companies like OpenAI and Google – with the slower, peer-reviewed publishing process in academia. Special attention is given to peer review and emerging publishing models such as Publish–Review–Curate (PRC). PRC addresses several shortcomings of the traditional peer review system and may provide a foundation for future adaptations aimed at accelerating scholarly publishing. Findings Given the rapid progress in both methodological and applied AI research, scientific publishing must adapt by accelerating its processes while preserving its core functions, particularly quality control. Originality/value Failing to keep pace with industry, academia risks becoming a second-tier player in AI development, trailing far behind industry research. This lag might seriously affect the direction and integrity of future science across all scientific fields and disciplines. Scholarly publishers may be sidelined – unless they adapt. Drawing on parallels with accelerated publishing during the COVID-19 pandemic, the paper calls on all stakeholders to act to ensure scholarly publishing remains vital in the AI age.