The Center for the Future of Trust builds tools, datasets, and analytical methods that help society understand how political ideas evolve — and how democratic reasoning can be strengthened in an era shaped by artificial intelligence.
We develop public-interest technologies that bring clarity to the information landscape. Our work supports journalists, policymakers, civil society leaders, students, and researchers who need transparent ways to see how narratives shift, where dissent forms, and how trust can be restored.
By Steph Buongiorno and Jo Guldi
Text Mining for Historical Analysis is under contract with Cambridge University Press, expected to be published in Spring 2026.
Text Mining for Historical Analysis is a practical guide for digital history on transforming large-scale textual corpora into interpretable evidence about the past using transparent, reproducible computational methods. It teaches readers how to clean, model, visualize, and interpret historical texts—from tokenization to embeddings to historiographic argumentation—while emphasizing methodology applied to the domain of history.
A benchmark and analytical method for mapping patterns of dissent and consensus across historical and digital corpora.
A no-code platform that visualizes how political speech changes over time, making complex data accessible to journalists, educators, and researchers.
An initiative that uses natural language processing to trace evolving narratives about environment, responsibility, and policy.
Marnie Hughes-Warrington, Distinguished Professor and Provost at the University of South Australia, gives her talk “When Will AI Win a Pulitzer Prize for History?” Overview: In July 2025, a report by a team at Microsoft research described a 91% overlap between the activities of historians and the tasks which Copilot completed to the satisfaction of users. Defenses of the distinctly human nature of historical methodologies and research followed. Has AI passed the Turing test for histories, the media asked? Should we call time and allow artificial historians to join historical associations, contribute journal papers, and enter and win history book prizes? This session argues for a broader view of the relationship between history and AI through the example of the Pulitzer prize for history. AI and history have a triple relationship: AI has a history; people use AI to make histories, and AI makes histories. Hughes Warrington will argue for a shift in thinking from the first to the third of these and for the broader recognition of how history making shapes AI. AI may win the Pulitzer Prize for History in future, but the nature of that history may be different.