AI Guiding Principles
At SDSU Library
ECAUG 2025, October 30
Artificial Intelligence Principles Task Force Charge
The Artificial Intelligence Principles Task Force will explore how AI can be responsibly and thoughtfully integrated into the SDSU Library’s work.
Through benchmarking, research, and environmental scans, the task force will deliver guiding principles and assessment criteria that center on ethics, transparency, equity, and sustainability in the adoption of AI technologies.
This document will help guide Library Administration’s decisions regarding the potential use of AI within the library and provide guidance on assessing any platform for use by the library.
Membership
Sarah Tribelhorn
Sciences and Sustainability Librarian
Keven Jeffery
Digital Technologies Librarian
Rayyon Robinson
Social Sciences and Data Librarian
Lucy Campbell
Electronic Resources Librarian
Connor Franklin Rey
Assessment Librarian
A Practical Problem
We are presented with AI tools by our content providers, which we can sometimes decide to turn on or off. How do we make these decisions?
Example: Primo Research Assistant
The AIM
This document provides guiding principles and a self-guided evaluative rubric to support ethical integration of AI tools already deemed to be useful into library workflows and services.
The SCOPE
In Scope
Library Workflows e.g., automation, scheduling, metadata, analytics
User Experience e.g., chatbots, research assistants, discovery assistants, accessibility tools
Information Literacy e.g., instruction, AI literacy, critical evaluation
Sustainability e.g., environmental, financial, and operational implications
Out of Scope
Personnel Matters e.g., staffing decisions, labor negotiations, hiring, RTP process
Timeline
Tuesday July 8
Tuesday
July 29
Thursday
September 11
Monday
October 20
Tuesday April 29
Monday
May 12
Tuesday
June 10
Tuesday
June 24
Kickoff Meeting Review Charge
Review Timeline�Review Mission / Vision / Values
Assign roles for research�Report - AI on campus
Share findings and recommendations from preliminary research
Brainstorm assessment criteria / draft rubric
Draft Guiding Principles
Library wide discussion of the draft principles - all staff and faculty
Review peer feedback Refine draft
Test draft criteria and guiding principles against an AI tool - refine
Share draft with SDSU Library Faculty for peer feedback
Review comments and notes from library-wide discussion.
Finalize draft to submit to Library Administration
HUMAN OVERSIGHT
The Guiding Principles
EXPLAINABILITY
ETHICS
ACCESSIBILITY
CRITICAL LITERACIES
SUSTAINABILITY
HUMAN OVERSIGHT
We champion human expertise - AI assists but does not replace library staff and faculty work. We encourage human review of AI outputs, support building expertise, and champion recognizing when to use it and when not to.
EMPLOYEES ARE MOTIVATED BY:
NOT GOOD (1) | MEH (2) | GOOD (3) |
Operates autonomously with no oversight. Replaces human decisions. It's not the best tool for the task at hand. | It has utility and capacity for human intervention is available but unclear or limited | Meets a specific need. Designed to support rather than replace human roles. Serves as a collaborator |
EXPLAINABILITY
We select tools that are understandable and provide documentation or training to ensure informed use. Where possible we will avoid vendor reliance by prioritizing open or in-house tools, and acknowledge these tools may rapidly change.
�
EMPLOYEES ARE MOTIVATED BY:
NOT GOOD (1) | MEH (2) | GOOD (3) |
Opaque operations, no documentation, proprietary tools locked down by vendors. The model is problematic | Provides basic explanation but very little transparency, training, or support. The model is unclear | Developed in-house or transparent user-friendly logic with training available. The model is easily identifiable |
ETHICS
We respect user autonomy, privacy, and intellectual freedom and support the option to opt out of adopting these tools. We clearly communicate when and how AI is used and we comply with institutional, CSU, and legal policies. We will not knowingly use exploitative models.
�
EMPLOYEES ARE MOTIVATED BY:
NOT GOOD (1) | MEH (2) | GOOD (3) |
AI is hidden. No consent or privacy protections in place. Negatively mitigates human experience. | Opt-out is unclear. Data use is unclear - e.g. how it is hosted / used/ shared. | Opt-out options clearly provided. Has a published privacy policy. |
ACCESSIBILITY
We commit to audit tools for bias and exclusion and promote inclusive equitable access to AI literacy by supporting underrepresented user communities in AI use
�
EMPLOYEES ARE MOTIVATED BY:
NOT GOOD (1) | MEH (2) | GOOD (3) |
Exclusionary or vendor siloed | May offer some accessibility features or meet basic accessibility guidelines, but also presents obstacles. | Designed and demonstrated to be accessible and inclusive |
CRITICAL LITERACIES
We promote AI literacy as an extension of information literacy and encourage users to develop personal research skills and maintain academic integrity.
�
EMPLOYEES ARE MOTIVATED BY:
NOT GOOD (1) | MEH (2) | GOOD (3) |
Encourages over reliance and discourages critical independent research | Neither encourages or discourages critique and/or independent research. | Encourages critical engagement (for example, cites sources) |
SUSTAINABILITY
We evaluate the environmental and financial impact of AI tools and choose interoperable, cost-effective, and maintainable solutions. We seek to minimize intensive use of AI when it has negative social and environmental impacts.
�
EMPLOYEES ARE MOTIVATED BY:
NOT GOOD (1) | MEH (2) | GOOD (3) |
High-cost resource. Reliance on vendors. Environmental impact is unclear | Long term environmental impact | Low-impact, cost-effective, supportive of SDSU strategic plan |
Example
Primo Research Assistant
Human Oversight: OK (2) - no opportunity for iterative adjustment
Explainability: OK (2) - no explanation of how search is being undertaken
Ethics: OK (2) - there is a privacy policy but it is unclear where the AI lives and if the data is being more broadly shared
Accessibility: Not Good (1) - Vendor Siloed
Critical Literacies: Not Good (1) - presents five articles as reputable without explaining why
Sustainability: Not Good (1) - Vendor reliant. Environmental impact is unclear
Total Points: 9/ 18
According to this rubric, SDSU Library would not adopt this AI Research Assistant Tool.
RESOURCES
Association of Research Libraries. (2024, April 25). Research Libraries Guiding Principles for Artificial Intelligence. Association of Research Libraries.
Association of College and Research Libraries. (2016). Framework for Information Literacy for Higher Education. In Standards, guidelines, and frameworks. American Library Association.
Digital Education Council. (2025). Digital Education Council AI Literacy Framework. Retrieved from https://www.digitaleducationcouncil.com/post/digital-education-council-ai-literacy-framework
Ex Libris. (2023). The impact of generative AI on libraries. Clarivate. Retrieved from https://exlibrisgroup.com/announcement/ex-libris-showcases-ai-innovations-shaping-the-future-of-libraries-in-2025/
Ex Libris. (2025). Ex Libris showcases AI innovations shaping the future of libraries in 2025. Retrieved from https://exlibrisgroup.com/announcement/ex-libris-showcases-ai-innovations-shaping-the-future-of-libraries-in-2025/
Guthrie, K. (2023). Approaching AI and advanced technologies together. JSTOR. Retrieved from https://about.jstor.org/blog/approaching-ai-and-advanced-technologies-together/
Hervieux, S., & Wheatley, A. (2024). Building an AI literacy framework: Perspectives from instruction librarians and current information literacy tools [PDF]. Choice360.org. Retrieved from https://www.choice360.org/wp-content/uploads/2024/08/TaylorFrancis_whitepaper_08.28.24_final.pdf
Kassorla, M., Georgieva, M., & Papini, A. (2024). AI literacy in teaching and learning: A durable framework for higher education. EDUCAUSE. https://www.educause.edu/content/2024/ai-literacy-in-teaching-and-learning/executive-summary
Lo, L. (2025). AI Literacy: A Guide for Academic Libraries. College & Research Libraries News, 86(3), 120. doi:https://doi.org/10.5860/crln.86.3.120
OpenAI. (2025, April 11). Cal State Tech Fridays – OpenAI ChatGPT [PDF]. California State University.
San Diego State University. (2024). Academic Applications of AI (AAAI) Micro-Credential. Retrieved from https://aaai.sdsu.edu/micro-credential
San Diego State University Library. (2025). AI at SDSU. Retrieved from https://libguides.sdsu.edu/AI/home
United Nations Educational, Scientific and Cultural Organization. (2024). AI competency framework for students. UNESCO Digital Library. Retrieved from https://www.unesco.org/en/articles/ai-competency-framework-students
Thank You