July 8, 2026
What can archives learn from AI? A new open-access article by Frances Corry, Assistant Professor in SCI’s Department of Information Culture and Data Stewardship, argues the answer is: quite a lot.
Published in Cambridge Forum on AI: Culture and Society, Corry's piece traces how "provenance”, the story of where something comes from, has been reshaped across archival studies and AI/machine learning. She shows how frameworks born in archival theory already shape how AI researchers document their datasets (think "Datasheets for Datasets"), then flips the question: what if that flow of ideas ran the other way? Could the transparency practices developed for AI training data help archives finally tell users the story behind a collection, not just its contents?
It's a sharp, timely case for cross-field thinking, and a reminder that the questions animating AI research and archival practice aren't as far apart as they seem.