MCN 2026

How can museums use AI to expand access to their collections without sacrificing the authenticity and trust that make those collections valuable? For the Computer History Museum (CHM), this question is not just strategic, it is fundamental to the institution's mission as a model for the ethical use of technology.

This session presents panel presentations and a prototype demonstration of CHM's ongoing work to use AI to expand online audiences while preserving institutional integrity. Panelists will include voices from the museum’s digital and curatorial teams and from external partners, offering a multi-stakeholder perspective on the opportunities and challenges of AI adoption. 

CHM has created one of the most significant oral history collections (https://www.youtube.com/playlist?list=PLQsxaNhYv8daKdGi7s85ubzbWdTB36-_q) in the technology world: hundreds of original long-form video interviews with the pioneers who built the modern computing industry. These recordings are invaluable to researchers but the length and format limit their reach with general audiences. CHM’s OpenCHM Digital Portal, launched in 2026, is designed to provide wide and deep access to CHM’s collections. Building on OpenCHM, an AI researcher and CHM collaborator based at Google DeepMind developed an AI-powered prototype that transforms this archive into a dynamic, navigable experience designed for the casual online visitor and the technology enthusiast. The prototype (https://f0lkl0r3.dev) was built using a system of LLMs and image generation models to process 875 long-form transcripts, extracting entities to create a sprawling, interactive knowledge graph featuring timelines, latent space exploration, and character mapping. The prototype creates an exciting new entry point but it also raises fundamental questions about longevity and trust. Without effective checks and balances, models can go off the rails (e.g. hallucinating facts about tech pioneers) and AI products created by third parties may not align with museum standards for accuracy. CHM’s oral histories are assets collected for the long term; they will inform the history of today when it is written decades from now. CHM must be a good steward of these assets, trusted by audiences and the subjects of current and future oral histories. 

The session will explore three interconnected themes. The creator of the prototype will offer ideas on what is possible for museums using AI. Members of CHM’s digital and curatorial teams will explore the strategic questions these opportunities raise for the museum: when AI makes so much possible, how do we decide how to invest our resources to ensure that AI is delivering real value? CHM will share its emerging guidelines for AI governance, risk management, and editorial workflows that may serve as a model for other museums. Our session will also include feedback from the audience on how their institutions are wrestling with similar challenges.