AI and the Ethical Implications for Libraries
an annotated biblography entry
Cox, A.M., & Mazumdar, S. (2022). Defining artificial intelligence for librarians. Journal of Librarianship and Information Science. 96100062211420–. https://doi.org/10.1177/09610006221142029
Summary
Cox and Mazumdar (2022) discuss artificial intelligence in relation to its practical use within the library field and the resulting ethical impacts it can carry for librarians. They provide a great starting point for readers by providing definitions of AI as found throughout research literature. The common themes within the definitions describe AI as using computers and algorithms to analyze user data to make decisions in processes normally done by humans. They follow up the section of definitions by providing five areas in which AI technology can be used within the library field. Cox and Mazumdar state, “They are areas in library back-end processes, in library services, through the creation of communities of data scientists, in data and AI literacy, and in user management” (2022). They also look at the potential impacts of AI in ethical and equity issues. The researchers state that AI is quickly being incorporated into mainstream society in search interfaces and the adoption of digital assistants like Google Assistant and Amazon Alexa. They even mention how AI is used for potential misinformation and fake news. The role of librarians in their tenet of navigating community members through information literacy becomes vital for this new state of information flow available through AI technologies. In addition, librarians need to be wary of how user data is protected and used with the possible incorporation of AI technologies.
Reflection
The paper provided a great introductory baseline of knowledge for anyone interested in AI within the library field especially considering initial searches for articles on the topic can result in some highly technical research papers. Although the authors of the paper have deep backgrounds in artificial intelligence and complex datasets, the paper is easily understandable with its well laid out structure and supporting tables. The source seems to successfully meet its goal of creating a baseline definition of AI within the library atmosphere.
The last example of how AI can be applicable in a library setting was particularly new to me. This example talked about how AI could not only be used to analyze a library’s collection as data, but also library user data. The ability to use AI technology to analyze the various statistics a library has on its own users for potential benefit seems very enticing. But because the authors mention that AI can also be biased, libraries and management must be very careful when using AI for potential prediction of library user behavior. Both library staff and users must be aware of any library privacy policies that explain the use, storage, and retention of individual user data. This opened up the fact that AI use in libraries is not just for collection management or access for staff and users, but also for the library in predicting community behaviors.