The STM Frankfurt Conference 2026 brought together more than 300 delegates from across the scholarly communications sector to discuss the implications of artificial intelligence (AI), research integrity, geopolitical uncertainty, and evolving research policies for scholarly publishing. Discussions focused on how publishers, researchers, librarians, and technology providers can work together to maintain trust in research and adapt publishing practices to changes in how scholarly content is produced, accessed, and evaluated.
Participants highlighted the growing role of AI agents in accessing scholarly content, rising research submission volumes, and increasing pressure on publishers to maintain quality and integrity. Despite these challenges, discussions reflected optimism about the sector’s ability to address emerging issues through collaboration, shared standards, and investment in research infrastructure.
Daan du Toit, Deputy Director General of South Africa’s Department of Science and Innovation, addressed the importance of international research collaboration amid geopolitical tensions and debates over scientific sovereignty. Drawing on the African concept of Ubuntu, which emphasizes shared responsibility, mutual trust, and collective progress, he argued that international cooperation remains essential to scientific advancement and sustainable development. He also cautioned against scientific decoupling and encouraged the research community to identify common ground when addressing global challenges.
The role of publishers in establishing standards and policies for trusted research in the AI era was another focus. As generative AI introduces new ways of discovering and accessing scholarly information, questions surrounding attribution, accountability, licensing, and discoverability are becoming increasingly important. A panel chaired by Violeta Ribarska of IOP Publishing, with Rose L’Huillier of Elsevier and Claudia Russo of STM, examined how the scholarly publishing community could contribute to discussions on regulation, technical standards, and infrastructure.
The panel drew on previous collaborative initiatives, including COUNTER, persistent identifiers, and the STM Integrity Hub, as examples of how the sector has addressed shared challenges. Participants emphasized that publishers need to remain involved in decisions that will influence how trusted research is accessed, attributed, and used.
A separate session introduced TRACE, a cross-community initiative intended to help ensure that trusted research remains identifiable, attributable, and measurable as AI increasingly mediates access to information. Chaired by Priya Madina of Springer Nature, the session featured Todd Toler of Ithaka S+R, Todd Carpenter of NISO, and Tasha Mellins-Cohen of COUNTER. Participants discussed the importance of enabling researchers to verify sources, ensuring appropriate recognition for authors and publishers, and helping libraries understand how scholarly content is being used.
Changes in China's research evaluation and publishing policies were also examined. A panel chaired by Nandita Quaderi of Clarivate, with Bob Howard of SAGE Publishing, Yu Sun of Tsinghua University Press, and Ting Yue of the National Science Library at the Chinese Academy of Sciences, discussed the country's shift from quantity-based research assessment toward greater emphasis on research quality, integrity, societal impact, and journal standards. The discussion covered investment in research infrastructure, publishing capacity, and journal development, alongside efforts to strengthen research integrity and domestic publishing platforms.
The conference also addressed the implications of AI for research workflows, quality assurance, and publishing accountability. A panel chaired by Rob Johnson of Research Consulting included Kim Eggleton of IOP Publishing, Maria Kowalczuk of Frontiers, Witold Kozakiewicz of the Medical University of Lodz, Indira Sen of the University of Mannheim, and Jonathan Woahn of Cashmere.
Participants considered how AI could support decision-making and reduce administrative workloads without compromising research integrity. Kim Eggleton emphasized the risk that increasing processing speed could undermine integrity and stressed the continued importance of reliable peer review. The panel agreed that AI could assist with publishing processes, but responsibility for editorial and publication decisions should remain with people. Jonathan Woahn described the current pace of change as a wartime moment, underscoring the need for organizations to respond quickly while maintaining appropriate oversight.
The closing leadership panel examined how publishers could respond to changing expectations around trust and technology. Chaired by Caroline Sutton of STM, the panel included Atul Goel of KnowledgeWorks Global, Laura Hassink of Elsevier/RELX, Steven Heffner of IEEE, Barbara Kline Pope of Johns Hopkins University Press, and Jessica Kowalski of Wiley. Participants noted that trust in scholarly content cannot depend solely on the reputation of a publisher or journal, making research verification and quality assurance increasingly important.
Panelists also discussed the potential for AI to redesign publishing workflows and allow professionals to focus on tasks requiring greater expertise. Rather than considering whether AI should replace peer review, the discussion centered on how the technology could reduce administrative work while preserving human judgment in editorial decisions. Steven Heffner highlighted the risk of failing to identify original and valuable research, noting the importance of human assessment in evaluating the novelty, significance, and scientific contribution of published work.
Organizational readiness and workforce development were additional areas of discussion. Participants highlighted the speed of technological change and the need for leaders to create environments that support responsible experimentation, skills development, and collaboration.
Click here to read the original press release.
