The AI Stack: Scaling Innovation from Policy to Collections

ID: WMA2026_584

Track:

How does a museum move from a “cool experiment” with AI to a sustainable workflow? This session explores Utah’s intentional steps to integrate Google Gemini across its workforce, providing a macro-to-micro blueprint. From state-level policy and departmental change management to practical collections applications like transcription, learn how to ASPIRE to innovate with a unified strategy. Includes a live prompting lab to translate policy into effective daily practice.

Session Information

Format: Regular session/panel (roundtable, single speaker, etc.)

Uniqueness: Bridges the gap between high-level government AI policy and grassroots museum collections work, offering a fully scalable blueprint for institutional AI adoption.

Objectives:

How do we thoughtfully integrate AI into museum practice without losing our institutional voice? This session outlines the intentional transformation of the State of Utah’s workforce via Google Gemini, offering a scalable blueprint for the museum field. Our objectives are to demystify generative AI governance and provide actionable workflows for museum staff. By examining this transition through three distinct lenses—State Policy, Departmental Strategy, and Collections Practice—attendees will understand how to safely utilize AI as a “copilot.” The session includes built-in discussion time and an interactive where attendees can see real-time solutions to complex transcription and metadata challenges. Expected Learning Outcomes:

1. Understand the Regulatory Landscape: Learn how frameworks like the Utah AI Policy Act create a “Safe Harbor” for innovation while protecting data privacy.
2. Identify Strategies for Adoption: Gain tools for departmental change management, including the creation of Standard Operating Procedures (SOPs) that build staff trust in AI.
3. Master Practical Workflows: Acquire immediately usable Google Gemini prompting techniques to clear transcription backlogs and enrich collection metadata for better public discoverability.

Engagement: We will conduct an “AI Triage” group activity. Attendees will review handouts (or slides) featuring various museum tasks (e.g., transcribing pioneer diaries, drafting grants, or generating exhibit labels for culturally sensitive items). In small groups, they will categorize these tasks into “Green Light” (safe for AI), “Yellow Light” (requires strict human-in-the-loop), or “Red Light” (AI prohibited) by applying the policy and practice frameworks discussed. We will conclude with a guided group debate.

Relationship to Theme:

Audience

Audiences: Curators/Scientists/Historians Registrars, Collections Managers Technology 

Professional Level: All levels 

Scalability: By breaking down AI implementation into three tiers (policy, department, individual practice), attendees from any size institution can adapt the framework. Small museums can utilize the boots-on-the-ground prompting strategies, while larger institutions can adapt the departmental SOPs and governance models.

Participants