Helping people and institutions make sound decisions with AI.
Open frameworks for AI literacy, institutional strategy, workforce development, and knowledge governance.
The work
Each one answers a different question, and they stack, from a single person typing a prompt out to institutions acting together.
Writing prompts that get useful results.
The CLEAR framework: concise, logical, explicit, adaptive, reflective.Published 2023It made prompting teachable rather than a knack, which is why it now appears in more than 200 institutional guides worldwide.
Lo, L. S. (2023). The CLEAR path: A framework for enhancing information literacy through prompt engineering. The Journal of Academic Librarianship, 49(4). doi:10.1016/j.acalib.2023.102720Judging an AI answer before you rely on it.
The CARE approach: classify, assess, review, enhance.Published 2026Review is the load-bearing step. It checks an answer against a record you actually trust rather than against your impression of it.
Lo, L. S. (2026). The CARE approach for academic librarians: From search first to answer first with generative AI. The Journal of Academic Librarianship. doi:10.1016/j.acalib.2025.103186Telling four kinds of AI answer apart, because each one fails differently.
The Answer Typography: factual, interpretive, constructive, strategic.Published 2026It does the classifying step inside CARE. Checking a factual answer the way you would check a strategic one catches neither.
Published with the CARE approach. doi:10.1016/j.acalib.2025.103186Setting up an agent that keeps acting after you stop watching.
The PASTAS checklist for context engineering. Six components, from purpose and audience through signals and review.Published 2026Prompting assumes you are in the conversation. An agent acts on conditions you set once and then keeps acting, so the work moves from writing a good prompt to engineering good context.
Lo, L. S. (2026). From prompt to agentic AI: The PASTAS checklist for context engineering. portal: Libraries and the Academy, 26(3), 403–415. doi:10.1353/pla.2026.a994552Moving on AI without waiting for certainty that is not coming.
Anchored Adaptation.Method and playbook available · Book forthcoming 2027Hold a stable capability framework, run small real experiments, capture what actually happened, then make a recurring decision to scale, revise, or stop. It exists so an organization can move without either freezing until the evidence arrives or committing to whatever launched last month.
Book: Anchored Adaptation: AI Literacy and Workforce Reskilling in Academic Libraries, ACRL, forthcoming fall 2027.Setting terms for what AI companies may do with your collections.
The UVA Archival AI Protocol.Published 2026 · v1.1A system that retrieves material and cites it can be stopped at any point. A system that absorbs it into model weights cannot, because nothing comes back out. Published under Creative Commons so other institutions can adopt it directly.
Lo, L. S. (2026). doi:10.18130/5dqf-9w86 · Adoption Kit · The reasoning behind itAdopting those terms alongside other institutions.
The Statement of Shared Practice.Live · 20+ signersA protocol one library adopts is a policy. The same terms adopted by many become something an AI company has to plan around. It launched in April 2026 with 12 founding institutions and has grown past 20.
library.virginia.edu/ai/shared-practiceWhat any adult should be able to do with AI.
The Universal AI Literacy Framework: five components, four levels, with rubrics.Preprint 2025The others are instruments for particular jobs. This one describes the underlying capability they all assume. Penn State and Clemson each credit it as a source in their own university-wide AI literacy frameworks.
Lo, L. S. (2025). AI literacy for all: A universal framework. Preprint, UNM Digital Repository, CC BY-NC 4.0. digitalrepository.unm.edu/ulls_fsp/213Who I am
I study how AI changes knowledge, judgment, and action, and how institutions can adapt under uncertainty.
I am Dean of Libraries and University Librarian at the University of Virginia and Special Advisor to the Provost on AI Literacy. I co-founded the UVA AI Literacy and Action Lab and co-chair the University's AI governance committee.
Before any of it, I trained as a screenwriter.