The Archives and Records Association UK & Ireland (ARA) has today published new open-access guidelines to help the Galleries, Libraries, Archives and Museums (GLAM) sector prepare their collections for the responsible use of artificial intelligence (AI).
Titled AI Preparedness Guidelines for Archivists, the four-page guidance has been produced through the ARA-funded FLAME project (AI For Libraries, Archives and Museums) and is freely available to the sector.
AI is now a growing topic of conversation across archives and record-keeping services, with managers and stakeholders increasingly asking whether it can help speed up cataloguing, identify sensitive content, or unlock new ways for users to access collections. The new guidelines aim to help organisations respond to these opportunities in a realistic, responsible and informed way.
The FLAME project is led by Loughborough’s Professor Lise Jaillant and Professor Giovanni Colavizza from the University of Copenhagen and the University of Bologna. It focuses specifically on the challenges and opportunities of using AI in the GLAM sector, with the guidelines forming the first major public output of the project. An open-access academic article related to the project will be published later this year.
The project aims to:
- Improve how GLAM organisations prepare digital collections before applying AI tools
- Support new professional practices around automation and AI in archives
- Help the sector move away from 'dark' archives to more accessible collections, while promoting more sustainable digital practices
The key message is clear: AI can support archival work, but only when collections are properly prepared. The guidelines stress that automation is not a magic solution, and that careful preparation, documentation and governance are essential to making collections 'AI-ready'.
The project was funded through the Archives and Records Association's Research, Development and Advocacy Fund, which supports research and innovation across the record-keeping profession. The FLAME project was one of two AI-related projects funded by ARA in September 2024, alongside a project led by Liverpool University's Centre for Archive Studies on AI and machine learning for catalogue conversion.