I've spoken to many librarians here in the US and across the world about the ins and outs of the library world. I've spoken to school librarians, public librarians, and academic librarians, as well as museum professionals. Across libraries and museum, AI is considered a major concern, for many reasons. Primary reasons why AI is a concern to information professionals typically have to do with copyright infringement, privacy concerns, misinformation, and environmental concerns. This makes sense, considering most of what information professionals study deal with these topics specifically.
By and large, I personally don't think libraries are responding to this problem in a particularly helpful way. Most people I've spoken to are disappointed in their leadership for how they handled AI. As a library, we can, and should, be doing a lot more.
Libraries in AI Survey
I took a poll on my Instagram account asking if their library had a policy regarding AI. 370 people responded and 71% (264) said that their library did not have a policy in place. This is alarming! Considering the fact that AI has many ethical concerns, including privacy and copyright issues, libraries really need to have a policy in place that will guide their staff in correct usage. Library Journal recently put out some guidelines on AI policy, and many of the suggestions it presents are reasonable and good to consider when drafting an AI policy for your library.
I ran a another poll on my Instagram story and Threads account asking basic questions surrounding AI in libraries. There were over 500 people who answered questions. Here were the questions and the responses to these questions:
Do you feel like your library is doing enough to address AI concerns?
YES (3%)
NO (78%)
UNSURE (19%)
Do you feel like your library leadership is adequately aware of ethical concerns surrounding generative AI?
YES (17%)
NO (61%)
MAYBE/UNSURE (22%)
Do you feel like your library leadership is primarily in favor or against generative AI?
IN FAVOR OF AI (45%)
AGAINST AI (16%)
UNSURE (39%)
Are YOU generally in favor or against generative AI?
IN FAVOR OF AI (1%)
AGAINST AI (94%)
UNSURE/NEED MORE INFORMATION (5%)
Do you have an MLIS or other library higher education?
YES (71%)
NO (21%)
NO, BUT I AM IN SCHOOL NOW (8%)
It is important to acknowledge that running a poll on social media is not the best way to get unbiased data as the "social media bubble" is a real phenomenon that cannot be understated. So, while this was not a perfect survey, 500+ respondents is still a decent size, and the poll number still confirmed my concerns. Librarians and library workers overwhelmingly feel that library leadership is not doing enough to address AI and the threats it presents to our profession and library leadership is not adequately aware of ethical concerns of AI and how it directly impacts libraries. Library leadership is also appears to be overwhelmingly in favor of AI, which contrasts how library workers with MLIS degrees feel about AI. Ideally library leadership should be primarily made up of people who have MLIS, but we all know that board members, administration, marketing, and HR can often not be library professionals.
Library Policy and AI
In addition to polling some of my followers, I also asked for examples of library policies that concern AI. In my opinion, some of these policies here are well written and cover the ethical nature of AI in a satisfactory way, but not all of these AI policies are adequate. Some of these policies are rather vague or actively are in favor of AI and do not adequately address serious AI concerns. I also reached out to my state Library Association president (a friend of mine and a fan of the page), and asked him if Arkansas had any public libraries with AI policies. He said he didn't know of any. As a profession, I think we are falling very far behind in this regard.
Some specific AI Policies that were sent to me:
DeKalb Public Library AI Policy
Hastings Public Library AI Policy
Houston Public Library System AI Policy
Waterloo Public Library AI Policy
Toronto Public Library AI Policy
Johnson County Public Library AI Policy
University of Groningen AI Policy
Marshall University Libguide
Additional information that was given to me regarding AI policies:
General AI training is mandatory in a Chicago library. According to this employee, it is very generic, but they have more specific training as well.
Orchestra librarian told me that because of super strict copyright laws, they are not allowed to use AI whatsoever.
Sno-Isle Libraries Organizational AI Guidelines says that AI can be used for drafting, refining text, and conducting research. AI should not be used for final drafts or direct customer interactions without human oversight. Overall their AI policy was very pro-AI.
One librarian said their library did not require the need to disclose the use of AI in any process unless it impacts the final outcome. The librarian in who sent this over to me said it was not a helpful policy and argued that every use of AI impacts the final outcome.
Many county libraries told me they basically offloaded their AI policy to the county government, rather than making their own policy.
Many libraries' AI policy focus on collection development (think, "primarily AI generated work is not allowed in library collections")
Looking Ahead
I think a real need in our profession is an increase in robust AI policies. For this reason, I plan on writing up an AI policy that can be adapted to libraries who wish to add an AI policy. Keep an eye out for that policy, coming soon.