The vast majority of scientific data fails to realize its complete capacity for sparking innovative breakthroughs.
For every 100 datasets generated in research settings, roughly 80 remain confined to their originating laboratories, 20 get shared but rarely see any meaningful reuse, less than two adhere to the FAIR guidelines, and typically just one contributes to fresh scientific insights.
These stark realities carry profound implications: they hinder advancements in cancer therapies, undermine the reliability of climate modeling due to insufficient supporting data, and contribute to the widespread issue of irreproducible studies.
In response to this crisis, the open-access publisher Frontiers has launched Frontiers FAIR² Data Management, heralded as the pioneering comprehensive service for research data powered by artificial intelligence. This innovative platform streamlines the entire lifecycle of data handling—encompassing curation, compliance verification, formatting optimized for AI applications, rigorous peer review, an engaging interactive portal, formal certification, and secure long-term hosting—into a unified, efficient workflow. Ultimately, it aims to transform current investments in research into accelerated progress across healthcare, environmental sustainability, and technological innovation.
FAIR² extends the foundational FAIR principles—Findable, Accessible, Interoperable, and Reusable—through an enhanced open framework that ensures all datasets are fully compatible with AI systems and ethically viable for reuse by humans and machines alike. As the inaugural practical realization of this advanced model, the FAIR² Data Management system emerges at a pivotal time when the volume of research outputs is exploding and AI is fundamentally altering the landscape of scientific discovery. It converts abstract guidelines into tangible, scalable infrastructure capable of delivering quantifiable benefits.
AI at the Core
Tasks that traditionally demanded months of painstaking manual labor—such as organizing datasets, validating their integrity, producing detailed metadata, and creating publication-ready materials—are now accomplished in mere minutes thanks to the AI Data Steward, which is driven by Senscience, the Frontiers initiative powering FAIR².
When researchers upload their data, they obtain four cohesive deliverables: a certified Data Package, a peer-reviewed Data Article that is fully citable, an Interactive Data Portal equipped with dynamic visualizations and AI-driven chat functionalities, and a prestigious FAIR² Certificate. Every component incorporates stringent quality assurance measures and concise summaries, rendering the data more comprehensible to non-experts and seamlessly integrable across diverse scientific fields.
Collectively, these deliverables guarantee that each dataset is safeguarded indefinitely, thoroughly validated, properly citeable, and primed for reuse, thereby expediting the pace of new discoveries while ensuring researchers receive the acknowledgment they merit. Moreover, Frontiers FAIR² boosts the discoverability and ease of access to these resources, fostering ethical reuse among scientists, policymakers, practitioners, community groups, and AI technologies, thereby maximizing the societal return on scientific funding.
Flagship Pilot Datasets
Frontiers FAIR² has been validated through several high-profile pilot datasets, each demonstrating its transformative potential:
- SARS-CoV-2 Variant Properties — This extensive collection examines 3,800 variants of the spike protein, integrating structural forecasts from AlphaFold2 and ESMFold with data on ACE2 receptor binding and expression levels. It serves as a vital tool for pandemic readiness, facilitating a more profound comprehension of how variants evolve, spread, and maintain fitness.
- Preclinical Brain Injury MRI — Comprising 343 diffusion MRI scans harmonized from four distinct research facilities, this dataset has been standardized across varying protocols to enable direct comparability. It facilitates the reliable identification of biomarkers, dependable analyses spanning multiple sites, and significant strides in the study of preclinical traumatic brain injuries.
- Environmental Pressure Indicators (1990-2050) — Merging historical observations with predictive models for 43 nations over a 60-year span, this resource monitors key metrics including emissions, waste generation, population growth, and GDP fluctuations. It forms the backbone for sustainability assessments and informed policymaking on climate issues.
- Indo-Pacific Atoll Biodiversity — Encompassing 280 atolls in five distinct regions, this dataset synthesizes records of biodiversity, reef ecosystem details, climate variables, and historical patterns of human utilization. It establishes a groundbreaking foundation for ecological simulations, prioritization of conservation efforts, and collaborative research on fragile island environments.
Feedback from researchers involved in these pilots underscores that Frontiers FAIR² goes beyond mere preservation and dissemination of data; it instills trust in its reusability via comprehensive quality validations, accessible summaries tailored for non-specialists, and the assurance of seamless integration across disciplinary boundaries—all while properly attributing credit to the originating scientists.
Every one of these pilot datasets fully conforms to the FAIR² Open Specification, ensuring they are meticulously curated, responsibly reusable, and dependable for sustained application by both humans and machines. This positions today’s data to propel tomorrow’s innovations addressing humanity’s greatest challenges.
Recognition and Reuse
Each instance of data reuse amplifies the inherent value of the source material, preventing any discovery from being squandered, empowering every contribution to ignite subsequent innovations, and providing researchers with the recognition their efforts deserve.
Dr. Sean Hill, co-founder and CEO of Senscience—the Frontiers AI venture driving FAIR² Data Management—emphasizes: “The scientific community pours billions into data generation, yet the bulk of it disappears without trace, and creators seldom receive due credit. Frontiers FAIR² changes that by making every dataset citable and every scientist acknowledged—truly honoring the vital labor of data production. This approach will hasten the delivery of cures, climate strategies, and emerging technologies to society, unleashing the full force of science.”
What Researchers Are Saying
Dr. Ángel Borja, Principal Researcher at AZTI, Marine Research, Basque Research and Technology Alliance (BRTA): “I highly recommend this approach to data curation and article publication, as it enables rapid generation of information in a format that’s practical for all types of end users.”
Erik Schultes, Senior Researcher at Leiden Academic Centre for Drug Research (LACDR) and FAIR Implementation Lead at GO FAIR Foundation: “Frontiers FAIR² captured the scientific aspects of the project perfectly.”
Femke Heddema, Researcher and Health Data Systems Innovation Manager at PharmAccess: “Frontiers FAIR² streamlines the application of FAIR principles for researchers and digital health professionals, demonstrating that rendering datasets like MomCare reusable need not be overly complicated. By promoting transparent, accessible, and practical data, it unlocks fresh possibilities in health research.”
Dr. Neil Harris, Professor in Residence, Department of Neurosurgery, Brain Injury Research Center, University of California, Los Angeles (UCLA): “Implementing Frontiers FAIR² offers an impartial evaluation of data completeness and quality across numerous dimensions. Such objective reviews and summaries enhance comprehension for those outside the field, ultimately boosting data sharing. As big data integrates into varied sub-disciplines, these validations and overviews will be essential for effectively utilizing and merging vast existing datasets in ongoing analyses.”
Maryann Martone, Chief Editor of Open Data Commons: “Frontiers FAIR² stands out as one of the simplest and most impactful methods to achieve FAIR data. Every principal investigator desires data that is findable, accessible, comparable, and reusable—within their lab, among collaborators, and throughout the broader scientific community. The primary obstacle has always been the substantial time and effort involved. Frontiers FAIR² substantially reduces that hurdle, making genuinely FAIR data attainable for the majority of research teams.”
Dr. Vincent Woon Kok Sin, Assistant Professor in the Carbon Neutrality and Climate Change Thrust, Society Hub, The Hong Kong University of Science and Technology (HKUST): “Frontiers FAIR² enhances the visibility and accessibility of our global waste dataset, aiding researchers around the world who frequently grapple with limited and disjointed data sources. I anticipate this will foster wider collaborations and speed up insights into sustainable waste management practices.”
Dr. Sebastian Steibl, Postdoctoral Researcher at Naturalis Biodiversity Center and the University of Auckland: “Achieving genuine data accessibility extends far beyond simply depositing spreadsheets into a repository. It involves rendering data intuitive to view, explore, and interpret without demanding extensive specialized training. The Frontiers FAIR² platform, featuring an AI chatbot alongside interactive visualization and summarization tools, renders our biodiversity and environmental data widely usable—not only by academics but also by practitioners, policymakers, and community-driven initiatives.”
Dr. Kamila Markram, co-founder and CEO of Frontiers, captures the essence: “Ninety percent of science vanishes into the void. With Frontiers FAIR² Data Management, no dataset and no discovery need ever be lost again—every contribution can now fuel progress, earn the credit it deserves, and unleash science.”
