A major concern library workers raise around AI use is the proliferation of large language models that can conflict with library values like user privacy and access to information. But alternatives exist to costly or proprietary, single models: Open-weight AI models that run locally on a workstation or laptop — and align well with the American Library Association Code of Ethics.
The latest post in our AI Tools for Academic Libraries blog series highlights some of the alternatives to familiar commercial frontier models such as ChatGPT or Claude, and how practical tasks like metadata generation and document analysis can be supported through open-weight AI models that more closely correspond to ethical principles.
Read the full post: AI Tools for Academic Libraries: Open-Weight AI Models for Local and Private Use
AI Tools for Academic Libraries is part of the OCUL AI and Machine Learning Initiative. The blog series is a collaboration between OCUL and Choice, a publishing unit of the Association of College and Research Libraries, and is hosted on Choice's LibTech Insights content channel that examines the day-to-day impact of library and education tech on academic librarians.