The dust is settling, the bubble has yet to burst and more libraries than ever have their AI policy in place, for users, for staff or hopefully for both. If you’re still working out where to start, we hope this post can help.
Many of these exist at an institutional level, where, for example, a university or local authority has a policy for AI use for all employees, including library staff. But we’re particularly interested in library-specific approaches, especially those that address the library’s role in supporting users’ AI skills and usage, as well as exploring their own engagement with AI tools. And as CILIP has correctly identified, librarians have a vital role to play when it comes to shaping AI policies.
Here are examples of some different policy approaches to the complex issue of AI in libraries, where both the uses of technology and its associated risks are constantly evolving.
National libraries have long led the way in using AI tools, so it is little surprise to see them implement formal policies that set out how and why they do so.
The National Library of Scotland has a detailed AI policy where it commits to a “curiosity and care” culture around AI, a lovely phrase for balancing the risks and rewards of AI technologies. This policy was created by a group of library staff that included representatives from metadata, development, digitisation, collections and research, maps, rights, and digital scholarship.
We are still very much early days when it comes to public library policies.
But Toronto Public Library’s Artificial Intelligence Policy has been a bit of a standard bearer for public libraries wanting comprehensive policies. And it really does strike a good balance – addressing privacy, transparency, sustainability, bias, and continual learning.
But this is for staff use of AI and the policy clearly states that the personal use of AI tools by library users is out of scope.
Subtitled the ‘Library at Queen’s University Belfast: Partnering on Responsible and Effective AI’, this Queen’s University Belfast Library Position Paper on AI really spells out the library’s important role in supporting responsible AI usage. As well as the six principles, this paper defines six roles for the Library AI Strategy Group and (my favourite bit) the specific workstreams that draw on the library’s expertise of “critical literacy, scholarly communication, rights/licensing, and long-term stewardship”.
When it comes to research libraries, many institutions have implemented an AI usage policy for staff and students. For example, the University of Oxford has produced extensive policies, but a key gateway to this is the Bodleian Libraries’ Using AI to Support Academic Work: A Library Guide: The University’s Policies on AI. This summarises and supports the broader AI policies of the university and helps link students to the relevant policies.
And while it’s more of a guide than a policy per se, Libraries Connected’s recent guide aims to help public libraries provide support for AI amongst their communities. And this is a useful perspective that really illustrates what libraries need to address in their own policies. Based on a project involving four UK library services, this approach focuses on boosting digital inclusion and AI confidence among the public, working with communities most at risk of digital exclusion. This reflects the public library’s role in helping communities understand and engage with AI safely and responsibly.
These are just a few different approaches to how libraries are leading on the responsible use of AI in their organisations and for their users. And it’s just a small sample of different kinds of libraries. We haven’t seen many published examples of school or business libraries, for example. But we’re keen to see more, so if you know of other library policies, please do share them with us.