Artificial Intelligence (AI) is a broad concept, with varying applications, rapid advancement, and constant changes. The Commonwealth Office of Technology defines AI as: “…the use of machine learning and related technologies that uses data to train statistical models for the purpose of enabling computer systems to perform tasks normally associated with human intelligence or perception." AI is not a new concept – it has been implemented in tools such as autonomous vehicles, GPS and mapping software, speech to text, gaming, productivity assistants, and more.
In November 2022, OpenAI released ChatGPT, an AI chatbot built upon a large language model (LLM). LLMs incorporate vast amounts of data during a training phase, with the purpose of using predictive algorithms to perform natural language processing tasks. While ChatGPT is not the only chatbot built upon an LLM, its release in 2022 sparked an ongoing, worldwide debate on artificial intelligence, its applicability, associated ethics, potential risks, and policy.
AI offers potential benefits to the information world – rapid retrieval, natural language interaction, data analysis at scale, productivity support, and more. It also presents numerous challenges for privacy, confidentiality, and accuracy, among other risk factors. Because of this precarious balance and widespread adoption, it is important for libraries to consider establishing policy guidelines and encourage or require staff training on appropriate use.
Integrated AI
Although many applications already utilized artificial intelligence in some capacity, the rise of large language models, LLM-powered chatbots, and generative AI has sparked a rush to incorporate these new advancements into as many sectors as possible. The library and information-related professions are not immune. Databases, online learning platforms, and information retrieval systems are integrating AI or experimenting with the technology. Archival services are working with AI-powered transcription, analysis, and tagging. And major search engines are incorporating AI synthesized responses to queries.
In work and productivity tools, such as word processing, spreadsheet, and email applications, options for AI assistance are becoming increasingly prevalent. PCs and other device manufacturers are deploying AI assistants and software at the operating system level. Even in marketing or design applications, built-in generative tools are being added.
Generative AI
Generative AI refers to applications with the capability of generating text, audio, images, video, and other information formats. To perform these tasks, an AI is trained on vast quantities of data to replicate and repurpose elements to fit the instructions it receives. Generative intelligence does offer beneficial uses, such as assistance with brainstorming, productivity, creation of marketing material, and grant writing. It also poses risks – invasion of privacy, potential copyright infringement, inaccuracy, deception, cybersecurity threats, and more.
Libraries must also contend with artificially generated content entering both physical and digital collections. An increasing number of AI-generated books and audiobooks are appearing on the market, with content entirely or partially generated by AI and attributed to nonexistent authors.
Policy Considerations
Artificial intelligence may be difficult to fully comprehend, especially due to the rapidity of development and deployment on a wide scale. The technology is constantly evolving, and a time investment in research and continuing education will be necessary to stay informed. As libraries consider responses to artificial intelligence, these are some starting questions to consider as policies are created or updated:
- Should the library include material fully or partially generated by AI in its collection?
- Are staff permitted to use AI tools, such as LLM chatbots?
- What guidelines does the library use when evaluating the integration of AI?
- What safety measures are in place to ensure confidentiality and privacy are maintained?
- What is the status of federal, state, or local legislation and regulation?
- Does the library actively encourage staff to stay up to date or seek out professional development related to artificial intelligence?
KDLA's Technology and State Data Coordinator and your Regional Library Consultant will be more than happy to assist you in exploring the needs of your library.