The MCLS management team has created a new MCLS Staff AI Use Policy that we will begin implementing in October, and I’m sharing information about it so our community is aware and can ask questions. 

We know AI comes with many questions (as studies from Pew Research Center, Gallup, and others have found) and it’s very important to us that we use AI in ways that foster trust and positive working relationships with libraries and our partners. MCLS is approaching using AI very intentionally and according to our guiding principles and documents (including our core values, strategic plan, and Board and internal policy).  

I’m using some questions I might expect libraries and our partners to have about how MCLS will use AI to frame this newsletter piece. I encourage you to contact me anytime with questions about AI (or anything else) at garrisons@mcls.org.  

Why did MCLS create an MCLS Staff AI Use Policy?
There are many reasons. First and foremost, as some MCLS staff have begun experimenting with using AI to search for, analyze, and brainstorm or draft information, we want staff interested in using AI to have clear expectations and guidelines so they do so ethically based on our core values and policies, and in ways that keep our internal data about libraries and our many partners safe from inappropriate exposure. Our relationships with libraries and partners are very important to us, and we do not want to use AI in any way that harms relationships or trust in us as an organization.

Second, AI functionality is popping up everywhere in software that libraries and MCLS use every day. We will use our policy as a framework to help us build ways to decide what AI competencies we expect our staff to have, how we’ll provide training, and which AI functionality we’ll allow staff to use in which software.

Third, library vendors have begun incorporating AI functionality into their products and including language about AI in their licenses. The more we understand AI and its implications, the better we can support libraries. 

How did MCLS create its AI policy?
We started by looking at other library consortia and libraries’ AI policies and asking AI to suggest policy elements we should consider including. One policy that was particularly helpful was the Kenosha Public Library Artificial Intelligence & Machine Learning Policy. Our management team reviewed multiple policy versions before approving the version 1.0 policy we are implementing now. Because AI changes rapidly, our management team will review and update our policy at least quarterly.  

How does MCLS plan to use AI?
As I noted above, we’ve already begun experimenting. If we use AI to help us create content that we make public, we will state that clearly. We will also explore how we might use AI internally to do rote tasks, data analysis, reporting and other work more easily and efficiently in ways that protect our internal data from being exposed outside closed AI that we and our software providers can control.  

How will MCLS protect my library’s data from inappropriate exposure?
Our policy is very clear that staff will not expose data that libraries and their staffs have entrusted to us to any publicly accessible AI system, for training or any other purpose.  

Will MCLS offer training on AI?
We have begun discussing this and will communicate about any AI training we may offer in the future.   

How will MCLS deal with ethical concerns around AI including but not limited to privacy, confidentiality, accuracy, bias, and AI’s resource consumption and environmental impact? 
Our policy includes a set of questions staff must ask before using AI, including how AI stores data and who has access, how we might mitigate any bias AI has based on its training data, and how we’ll verify that what AI generates is accurate. It also has sections on data privacy and confidentiality, and we state clear expectations including that staff must not use AI to generate content that is unethical, offensive, or in violation of MCLS policy. While we see great potential to gain efficiency through AI, we are conflicted about the fact that AI consumes a great deal of computing power, water required to cool servers, and other resources. We also have concerns over regulations (or the lack thereof) regarding AI in our current volatile times. 

Will you share your MCLS Staff AI Use Policy?
No, not at this time. We are in the process of implementing the policy with our staff and have more work to do to ensure we have clear expectations and processes to follow. We want you to be aware that we have a policy, for the reasons outlined above. You are welcome to ask me any questions you may have about our policy, at garrisons@mcls.org or by phone at 517-489-4071.  

What AI competencies does your library consider most important for staff? What AI training have you or your staff have taken that you’ve found helpful? What questions or concerns might you have about how MCLS might use AI to improve how we serve your library? What else might be on your mind regarding AI, or anything else?

Let me know at garrisons@mcls.org.