Abstract
Curated Information Frameworks (CIF) are models for human-centered AI governance that rethink transparency not as maximal disclosure but as a curated, participatory process. Current debates often falsely frame transparency as a binary: either AI systems should be fully “opened” to scrutiny, or they will remain opaque “black boxes.” Both approaches fail to capture the necessary reality and nuance of real-world systems. Excessive technical detail overwhelms users, while opacity may undermine trust. My CIF offers a third way: contextual, role-sensitive curation of information that aligns with user expertise, local contexts, and ethical stakes. Grounded in John Dewey’s philosophy of democratic inquiry, my CIF reframes transparency as an iterative, community-embedded process rather than a release of facts. Dewey argued that knowledge becomes meaningful only when situated within lived experience and shared inquiry; CIFs operationalize this by co-developing explanatory frameworks with affected stakeholders, ensuring that AI systems communicate in ways that are actionable, comprehensible, and just. I illustrate this approach through a 2024 case study, conducted in conjunction with the U.S. National AI Research Institute for Agricultural AI (AgAID). In participatory workshops on an autonomous apple picker, farmworkers, managers, and researchers collaboratively constructed a CIF: a tiered, role-specific transparency map clarifying who needs to know what, when, and how. Rather than being inundated with uninterpretable data, orchard workers received contextual explanations about machine reliability, while regulators gained insight into governance mechanisms. This process not only fostered trust, but also redefined key concepts and gave ownership of the system’s information to agricultural communities.
Presenters
Emily La RosaPhD Candidate, Philosophy, Michigan State University, Michigan, United States
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
2026 Special Focus—Human-Centered AI Transformations
KEYWORDS
AI, Human-Centered, Governance, Curated Information Framework, Transparency, Reliability, Ethics, Agriculture
