From FAIR to AI-FAIR Metadata-First Infrastructure for AI-Enabled Science

Abstract

Artificial intelligence (AI) technologies and their capabilities have advanced rapidly over the past few years, leading to widespread interest and adoption in the biological and ecological sciences. However, the data infrastructure and computational resources that provide the foundation for AI workflows in ecology and biodiversity science remain fragmented. Moreover, the flow of value remains largely unidirectional: data move from collectors, through repositories and aggregators into models, and finally to research products or management actions. The derived value, such as corrected taxonomies, annotations, quality flags, and insights, rarely returns to source communities. We propose an extension of the original FAIR principles to address the needs of AI-ready infrastructure: the AI-FAIR metadata criteria, which place AI-enabled science in the context of a bidirectional Question-Data-Model triangle. This framework is designed to facilitate large-scale content-based analysis across diverse collections and ensure that the scientists generating the data benefit from the AI-enabled advances built on their data. Alongside these criteria, we introduce an agentic evaluation workflow that both rates existing datasets on the AI-FAIR and traditional FAIR criteria and provides actionable feedback where critical metadata is missing or incomplete. This modular dataset evaluation tool, paired with a domain-appropriate checklist, translates these principles into a reusable assessment of machine-learning readiness, AI-readiness for specified tasks, traceability, and ethical and CARE-aligned stewardship. Through this approach we aim to create a community-informed system to close the model data loop, so that AI also enriches data to further enable scientific discovery.

From APTs to advanced persistent biological threats (APBTs): a taxonomy and large-scale metadata audit of genomic surveillance infrastructure
N. Anjum, M. Kiran, H. A. Alfraihi +3
bioRxiv9 Oct 2026PreprintCC BYFull textGenetics & genomicsbioRxiv
Verifiable Agentic Science on Public Infrastructure
D. Baker, S. Cain, M. van den Beek +10
bioRxiv26 Oct 2026PreprintCC BYFull textBioinformatics & computational biologybioRxiv
Expanded implementation of Fast & Fair paid peer review reduces time to first decision without reducing review quality in a biology journal
D. A. Gorelick, A. Clark
bioRxiv2 Jun 2026PreprintCC BY-NCFull text3 citedGeneral biologybioRxiv
EMMA: an R/Bioconductor package to automate tracking of metadata in functional enrichment analyses
Najla Abassi, Annekathrin Silvia Nedwed, Federico Marini
arXiv24 Sep 2026PreprintCC BY 4.0Full textGenetics & genomicsarXiv
The Shape of Biological Metadata: Measuring Repository Richness with Entity-Based NLP Metrics
M. J. Rodriguez-Cubillos, T. Zielinski, J. R. Swedlow +2
bioRxiv8 Sep 2026PreprintCC BYFull textBioinformatics & computational biologybioRxiv

Cite

BibTeX RIS

BibTeX

@misc{campolongo2026from,
  title = {{From FAIR to AI-FAIR Metadata-First Infrastructure for AI-Enabled Science}},
  author = {Elizabeth G. Campolongo and Eric R. Sokol and Marty Downs and Gina Glenne and Gregory E. Maurer and Sydne Record and Dustin R. Rubenstein and Matthew J. Thompson and Christine Laney and Hilmar Lapp and Max Tarjan and Tanya Berger-Wolf},
  howpublished = {arXiv},
  year = {2026},
  month = nov,
  doi = {10.48550/arxiv.2611.03327},
  eprint = {2611.03327},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2611.03327},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - From FAIR to AI-FAIR Metadata-First Infrastructure for AI-Enabled Science
AU  - Elizabeth G. Campolongo
AU  - Eric R. Sokol
AU  - Marty Downs
AU  - Gina Glenne
AU  - Gregory E. Maurer
AU  - Sydne Record
AU  - Dustin R. Rubenstein
AU  - Matthew J. Thompson
AU  - Christine Laney
AU  - Hilmar Lapp
AU  - Max Tarjan
AU  - Tanya Berger-Wolf
PY  - 2026
DA  - 2026/11/02
PB  - arXiv
DO  - 10.48550/arxiv.2611.03327
UR  - https://arxiv.org/abs/2611.03327
AB  - Artificial intelligence (AI) technologies and their capabilities have advanced rapidly over the past few years, leading to widespread interest and adoption in the biological and ecological sciences. However, the data infrastructure and computational resources that provide the foundation for AI workflows in ecology and biodiversity science remain fragmented. Moreover, the flow of value remains largely unidirectional: data move from collectors, through repositories and aggregators into models, and finally to research products or management actions. The derived value, such as corrected taxonomies, annotations, quality flags, and insights, rarely returns to source communities. We propose an extension of the original FAIR principles to address the needs of AI-ready infrastructure: the AI-FAIR metadata criteria, which place AI-enabled science in the context of a bidirectional Question-Data-Model triangle. This framework is designed to facilitate large-scale content-based analysis across diverse collections and ensure that the scientists generating the data benefit from the AI-enabled advances built on their data. Alongside these criteria, we introduce an agentic evaluation workflow that both rates existing datasets on the AI-FAIR and traditional FAIR criteria and provides actionable feedback where critical metadata is missing or incomplete. This modular dataset evaluation tool, paired with a domain-appropriate checklist, translates these principles into a reusable assessment of machine-learning readiness, AI-readiness for specified tasks, traceability, and ethical and CARE-aligned stewardship. Through this approach we aim to create a community-informed system to close the model data loop, so that AI also enriches data to further enable scientific discovery.
ER  -

Abstract

Related

Cite