This Statement sets out the approach of the National Library of Scotland to the responsible adoption and use of Artificial Intelligence. It will guide our decision-making on how we employ AI tools, communicate about our use of AI, collect AI-generated material, and allow our data to be used by others.
Artificial Intelligence (AI) is a broad term used to refer to a range of algorithm-based technologies capable of performing tasks commonly requiring human intelligence. For the purposes of this Statement, we will use “AI” to refer to computer-driven automation systems with the potential to impact the people and work of the National Library of Scotland.
AI applications such as image and text recognition have been in use for a long time, whilst other areas are developing at a fast pace – for example, Large Language Models (LLMs) and generative capabilities around text, imagery, audio and video.
Without familiar patterns of use, established standards and regulation, certain types of automation can present risks of misuse and harm dependent on their context, scale and application. These include data protection issues, copyright infringement, digital exclusion, patterns of bias, environmental impacts and further ethical considerations.
At the same time, we recognise that AI tools present multiple opportunities for libraries to improve how we work including improving access to content, automating processes around the management of collections and enabling new forms of research and creativity with collections.
As a National Library we have a role to play as a trustworthy source of information and a responsible steward of collections and data. We have both an immediate and generational responsibility to prioritise both our people and collections, providing reliable access to knowledge and information, and creating spaces for debate and discussion. This requires us to think carefully about how and when new technology is used.
Our commitments
This document was created with reference to and awareness of a range of AI definitions, statements, and policies from sources we trust. These include:
Last revised: 24 May 2024.
[1] See: John Fitzgerald (SCVO), Careful now! Your emerging AI Strategy, 19 October 2023
[2] Early examples of AI generated material already in the collection include published books, magazines, web pages, template-driven content inside newspapers such as weather report and sudoku, and visual effects in films.