AAAS and DataSeer test AI automation for research reporting checklists
The American Association for the Advancement of Science will pilot automated generation of MDAR checklists-documents that verify materials, design, analysis, and reporting details in manuscripts-with the software company DataSeer over the next six months.
MDAR checklists support transparency and reproducibility in scientific publishing but require significant manual work from authors and editorial staff. The pilot will test whether AI can produce these reports directly from submitted manuscripts without human intervention.
DataSeer will deploy its SnapShot technology to generate pre-filled MDAR reports based on manuscript content. Science journal editors will then review the automated reports to assess accuracy, usefulness, and whether they fit existing editorial workflows.
Valda Vinson, Executive Editor of the Science journals, said the collaboration addresses a core challenge: "Ensuring clear and consistent reporting is central to the integrity of the scientific record. This pilot will help us understand how AI-driven tools can support our editors and authors in meeting these expectations more efficiently."
The partnership reflects a practical question facing editorial teams: can automation reduce repetitive work while maintaining standards? Success could inform how other publishers approach research transparency workflows.
Tim Vines, Founder and CEO of DataSeer, said the pilot will evaluate "how automated MDAR generation can support editorial teams while preserving the high standards expected of Science."
Results may also provide insights into how automated workflows could strengthen data sharing and reproducibility across scholarly publishing more broadly.
For those working in research administration or editorial roles, understanding these emerging workflows is increasingly relevant. Consider exploring AI Agents & Automation to understand how these tools function in practice.
Your membership also unlocks:
