科研与出版领域的负责任创新研究报告

In this report Foreword3 Executive summary The quiet revolution: the rise of AI in peer review Surface use, deep potential9 Unlocking AI’s potential globally and across generations11 Bridging the gap to impactful AI use in research13The trust gap13The training gap16The policy and governance gap18 20 For educators and research institutions20For funders and policy makers21For tool developers and providers21For publishers21 Frontiers’ framework for responsible AI governance23 Conclusion26 Appendix27 Foreword Frontiers was established to strengthen peer review – the foundationof scientific trust – by combining openness, collaboration, andtechnological innovation. Today, that same vision guides howwe approach artificial intelligence. AI holds immense potential to transform scientific research, and its responsible integrationstands among the field’s greatest opportunities to help publishers deliver rigorous, fair, andefficient peer review. AI has already redefined the authoring process for scientific research, but it is now quietlyreshaping how it is reviewed. Our 2025 global survey reveals that while AI is steadily finding itsplace in peer review, its full potential is currently untapped. Today, it is often used for surfacetasks, like polishing language, drafting text, or handling administration, rather than for deeperanalytical and methodological work where it could truly elevate rigor, reproducibility, andscientific discovery. At the same time, the unethical use of AI to mass-produce low-quality or fabricated researchposes new challenges for peer review. But these risks only sharpen the imperative to harness AIas a force to combat fraud; one that scales quality and strengthens integrity across the entireresearch cycle. With the right leadership and guardrails, AI can act as an elevator for researchquality, rather than a shortcut around it. This study, the first of its kind, captures the realities of AI adoption, researchers’ perceptions, andthe challenges shaping its integration in research. Unlike previous surveys, we also measure whereand how researchers seek AI training and identify critical governance shortfalls across institutionsand publishers. Frontiers has long been a pioneer in using artificial intelligence to support academic publishingat scale. Our unique three-pronged approach to quality combines the human expertise of oureditorial boards, our in-house teams, and our own AI tools. Our AI assistant (AIRA) supports reviewby performing quality checks that would be impossible at scale through human effort alone – butalways in conjunction with human validation. We’ve published policies for authors, editors, andreviewers that encourage responsible and transparent use of generative AI technologiesthroughout the publishing process. Drawing on our own AI journey and informed by insights from the global research community,this report is our call to action: a catalyst for responsible transformation. It explores whatstakeholders across the entire research ecosystem – from funders to individual reviewers– can do to guide innovative AI usage so that it translates into impact. We’re excited to leadby example on this journey toward a future where rigorously reviewed, open research is trulyenhanced by AI and helps society find the solutions it needs for healthy lives on a healthy planet. Kamila Markram CEO and co-founder,Frontiers Executive summaryKey findings The quiet revolutionThe rise of AI in peer review Researchers are ahead of governance Global and generational insights informhow to unlock AI’s potential in research Call to action: turning insight into impact Closing the trust gap can increase confidence Methodology The quiet revolutionThe rise of AI in peer review While AI’s impact on research writing has been widely reported and impossible to ignore, itsquiet rise in peer review tells a subtler story – one of steady normalization and experimentation,but enormous untapped potential. Harnessing AI purposefully is essential to shaping the industry’s conversation about the futureof peer review, a process long challenged by issues of sustainability, transparency, and fairness.Used responsibly, AI can ease these pressures and reinforce trust in the evaluation of science.Peer review is where the next wave of meaningful progress will come from. Our findings show that more than half of peer reviewers (53%) now use AI, markinga profound cultural shift from traditional peer review to one that increasinglyintegrates AI tools. Q: How often do you use AI tools when peer reviewing articles? Adoption trends reveal that nearly one in four reviewers (24%) have increased their AI usein the past year, signaling progressive normalization. Regular use remains limited, though– a reminder that normalization is only the first step toward confident adoption. I had never used AI to help with peerreview until the last four months orso. It has substantially cut down ontime needed for each review. As