The 4th Sino-Swiss Research Integrity Workshop was jointly hosted by Swissnex in China, the Embassy of Switzerland in China and Frontiers recently. As Switzerland and China are celebrating the 10th anniversary of the Innovative Strategic Partnership in 2026, by bringing together speakers from leading institutions in both countries, the workshop moves from frameworks to best practices, seeking to advance a shared Sino-Swiss contribution to a fure-ready global research ecosystem on which the world can rely.

On behalf of the organizer,Frederick FENTER, Chief Executive Editor of Frontiers， addressed the welcome Remarks.Frontiers was founded in 2007 by Henry Markram and Kamila Markram, neuroscientists from the Swiss Federal Institute of Technology (EPFL) in Lausanne, Switzerland.It is one of the world’s largest and most impactful research publishers, dedicated to making peer-reviewed, quality-certified science openly accessible.

How research integrity must continually adapt to a research ecosystem increasingly shaped by artificial intelligence? What roles can—and should—artificial intelligence play in research and publication?Recently,Global People had a conversation with Frederick FENTER.

Global People：Has AI had a significant impact on your team's work over the past period? Which specific areas of your work have been affected by AI?

Frederick：Frontiers uses AI across the entire editorial process. We have a product called AIRA — the Artificial Intelligence Review Assistant — which we released as early as 2018. It performs over 40 standardized checks and quality assessments, including image manipulation detection, citation anomaly detection, and making sure that authors have followed all ethical guidelines.

AI advances so quickly. Authors are using artificial intelligence more and more often to prepare their articles for publication. Reviewers are already using AI. We have to understand best practices almost in real time as AI advances.

Our answer at Frontiers is to provide roadmaps rather than telling people what they cannot do. We published the AI Playbook in April 2026. This is an important resource that covers the full publication lifecycle for all the people involved, depending upon their role. This is very much a living document because the technology evolves so rapidly.

Global People：What is your views on China's development in AI technology and digital media applications? Any potential for future cooperation ?

Frederick：My prediction is that if we pull together, if we're able to share insights and experiences, the challenges in research integrity are ones we will be able to put behind us. We also take a lead in this publishing industry to share our insights and our experiences with our colleagues.

The Sino-Swiss Research Integrity Workshop has been held since 2023 and is now in its fourth edition, co-hosted by Frontiers, the Swiss Embassy in China, and the National Science Library of the Chinese Academy of Sciences. It serves as an official platform for exchange on research integrity practices between China and Switzerland. Over the past years, we have invited scholars from institutions including China's Ministry of Science and Technology, the Swiss Academy of Sciences, Tsinghua University, Peking University, the University of Zurich, and ETH Zurich for in-depth discussions.

In order for this to work well, we have to mandate transparency — so people declare in which contexts they're using artificial intelligence. We have to embed artificial intelligence literacy inside the research system. We have to strengthen integrity and oversight. And we have to make sure that access to artificial intelligence is equitable. All of these require cross-border dialogue and collaboration.

Global People：Will AI exacerbate existing inequalities in the global flow of information?

Frederick：This risk is real. An open access publisher like Frontiers does not look to license its content exclusively to AI companies for tens of millions of dollars. Other publishers do this because they consider the content to be their intellectual property, even though it's been paid for by public funding. It really belongs to the public.

The critical difference is that in open access channels like Frontiers, articles are published under a Creative Commons license, the CC BY agreement, that fundamentally allows AI algorithms to use and mine that content, provided there is an adequate and proper attribution. If you give attribution, you've satisfied the terms of the license.

In order for artificial intelligence to reach its full potential, we have to make sure that access to artificial intelligence is equitable.

Global People：Has AI made you and your team's work more efficient?

Frederick：All of the submitted manuscripts are put through AIRA. It does 40 standardized checks and quality assessments. After this very first stage, about one third of the articles are already rejected — removed and not sent to peer review. The humans have decided, using AIRA, that these articles are not of sufficient quality to even go to peer review.

There are 1,000 questions that can be asked to make sure that the overall picture, the overall foundational structure of this contribution makes sense. That's something that AI actually could be very good at.

In many ways, AI can almost become better than humans. But the humans always decide. Humans are the ones that take responsibility for all the critical decision points.