As libraries define their relationship with artificial intelligence (AI), library policy will play an essential role. History tells us that technological developments have a tendency to thwart well-intentioned policy, and this may prove especially true with respect to AI!
It’s Always the Time for Policy Review
In order for policies to remain effective and relevant, it’s important to review them from time to time. In some cases, such review may be necessary–in New York, library minimum standards require that public libraries review all of their policies every five years. Regardless of any legal requirement, such reviews should be completed as a continuous, incremental process, though circumstances may dictate which policies are given priority.
Artificial Intelligence is the Elephant in the Room
As you engage in your regular policy review, you should do so aware of the capabilities of artificial intelligence, generative AI in particular. This awareness can come from hands-on experience, the experience of our colleagues, and engagement with AI discourse and current events. Consider the ways AI might place stress on that policy.
For example, imagine you are reviewing your library’s reconsideration of materials policy. Typically, such a policy requires that a person objecting to a physical or digital piece of library media reads, views, or otherwise engages with the item. They then perform some intellectual labor, such as filling out a form stating their objections. Then there’s the labor performed by staff; they read the form, examine the material in question, and perform their own review. This can require significant time and money; the Nazareth Area School District in Pennsylvania spent $109,931.74 to pay staffers to review 23 challenged books!
Now let’s consider the AI factor. We know generative AI’s capability as a productivity tool/writing assistant. Imagine a single person using ChatGPT to write 50 challenges over the course of a weekend! Does your policy require you to review all of them? How quickly? In this scenario, a common AI tool is used to break a widely-held library policy.
Policy Reviews Need Not Acknowledge AI to Address the Technology
Looking at the reconsideration of library materials policy, our instinct might be to try and prohibit AI’s use in authoring a challenge. While this is understandable, AI detection tools can be unreliable, biased, and there are “humanizing” AI tools being built to actively thwart them! A revised reconsideration policy that places common-sense limits on the number of challenges a person can submit over a period of time could mitigate generative AI’s impact, without ever explicitly mentioning the technology itself.
Do I Need AI-Specific Policy?
The short answer is yes! Policy can help you set guardrails–this is especially important when dealing with a widely unregulated technology! Right now, library staff are using generative AI, likely without any official guidance on what is organizationally appropriate. For that reason, I recommend prioritizing an AI tool use policy for staff. Such a policy should address several points, including:
- A definition of artificial intelligence: AI means different things to different people. Defining the technology is an important preamble to a policy document that attempts to regulate its use!
- Data security: no sharing of logins.
- Privacy: personally identifiable information should not be entered into AI tools, nor data that is not publicly available.
- Using high-quality, vetted/approved tools: it’s best not to use some fly by night app of dubious origin. The library can maintain a vetted list, which should be subject to its own review–a tool’s utility or terms of service may not always reflect library values or best practices!
- Approved or disapproved use cases (ex: should a person use an entirely AI-generated piece of writing with no human intervention?) Is disclosure/attribution required?
For some examples of recently crafted AI usage policies, have a look at those of the Hastings Public Library in Michigan, and the Kenosha Public Library in Wisconsin.
Until Next Time!
At present, the age of AI can feel like the wild west! In order to successfully navigate this uncertain environment, we’ll need to craft thoughtful, relevant policy that is either AI-specific or AI-aware. Let’s get to it!
Items of (Potential) Interest
- I was interview by Library Journal for the article AI in the News, which explores the impact of generative AI on the news media, and media literacy.
- I’ll be providing a keynote at the Generative AI in Libraries Annual Conference, which takes places on June 9-12, 2025. Registration should be opening soon.
- I’ll be at ALA 2025 this year, part of a very talented panel on “The Future of Library Services: Innovating with AI, User Experience, and Service Design”. The preliminary schedule is currently available.
- I was a guest on the Patchogue Medford Library podcast, Adventures in Parenting, where I spoke about AI and our youth.
As always, if you’re looking for a speaker for your library event or staff development day, let’s talk! I cover the intersection of emerging technologies and libraries, keeping it conversational, informative, and entertaining.