Smithsonian Applied AI to Revolution Era Artifacts

The institution is using machine learning to link historical items across its vast collections for a 2026 project.

Updated on Oct. 3, 2026 in Artificial Intelligence

Isometric editorial illustration showing an antique brass compass and a historical wax seal connected by a thin geometric line. — The Smithsonian Institution is applying machine learning to link 10,000 artifacts from 1770 to 1810, aiming to uncover new historical connections. AI Illustration. Upload story photo >

Live Poll

Do you support the use of artificial intelligence to manage and interpret historical records?

In the spring of 2026, the Smithsonian Institution began applying AI models to digitize and connect approximately 10,000 artifacts from the 1770 to 1810 period. This initiative, known as the Revolution Crossroads project, seeks to surface new historical narratives from the nation's founding era.

Why it matters

By automating the cross-referencing of catalogs from the Smithsonian and the Library of Congress, researchers can identify connections between disparate objects that were previously difficult to track. This approach enables a more comprehensive public record of individuals living during the American Revolution.

The project analyzes a 40-year window of artifacts using machine learning models to search across massive institutional databases. These models compare digitized object data against Library of Congress records to uncover associations like the recent link between a specific silversmith and a creamer.

The players

Smithsonian Institution

An organization operating 21 museums and galleries that is now applying AI to its historical catalogs.

Library of Congress

A research library that serves as a primary source for external data in the institution's cross-referencing project.

National Museum of American History

A Smithsonian-operated museum where researchers recently used the technology to link a silversmith to a specific creamer.

The details

Researchers first digitized or redigitized objects from the 1770 to 1810 era to provide the high-quality input needed for machine learning models. These algorithms then scan and match metadata across the Smithsonian and the Library of Congress, surfacing potential links between items. Historians provide necessary oversight to interpret the AI-generated results and identify nuances that automated systems might misinterpret.

1770-1810: The period of history covered by the identified objects.
1774: A silversmith placed newspaper advertisements for a creamer later linked to the project.
1846: The year the Smithsonian Institution was founded.
Spring 2026: Researchers began applying AI models to the Revolution Crossroads project.

The Tech Race

The project aligns its technological deployment with the upcoming 250th anniversary of United States independence to expand public access to founding-era history. It mirrors broader efforts to modernize humanities research by applying computational analysis to massive, previously siloed institutional archives.

The public can expect to see new information added to the historical record about individuals from the nation's founding period as the project surfaces new connections. These findings will become accessible to scholars and the general public, providing a more detailed look at life in the late 18th century.

The takeaway

The Smithsonian is demonstrating how large-scale digitization and AI can transform static archival holdings into dynamic, connected historical networks. Readers should watch for future updates as the institution releases new data findings leading up to the 250th anniversary of independence.