AI Deployed to Locate Missing U.S. War Veterans

The POW Investigative Project is using artificial intelligence to analyze records of more than 9,000 missing Americans.

Updated on Sept. 18, 2026 in Artificial Intelligence

Isometric editorial illustration featuring archival folders and geometric data blocks, representing the digital analysis of historical war records. — The POW Investigative Project is utilizing artificial intelligence to process over one million declassified records, aiming to locate more than 9,000 U.S. service members missing since 1951. AI Illustration. Upload story photo >

Live Poll

Should the government prioritize using new AI technology to resolve past war-era missing persons cases?

The POW Investigative Project has begun using artificial intelligence to process massive archives of declassified documents, satellite imagery, and foreign-language data. This technology aims to locate over 9,000 American service members who remain unaccounted for from the Korean, Vietnam, and Cold Wars.

Why it matters

By automating the analysis of 50,000 project files and 1 million Senate-declassified documents, the group hopes to overcome the significant hurdles in historical evidence recovery. This shift aims to improve the efficiency and reach of searches that previously faced challenges in foreign archives and linguistic barriers.

The project uses AI to synthesize 1 million Senate-released documents and 50,000 files, including satellite maps and foreign-language internet archives. The technology also restores faded records and renders black-and-white sketches into color to assist identification.

The players

POW Investigative Project

This group researches the status of American troops missing from the Korean, Vietnam, and Cold Wars.

Senate Select Committee on POW/MIA Affairs

This body released 1 million declassified records in the 1990s that the current project is now utilizing.

Alexander Bonnyman Jr.

A service member whose remains were identified in 2015 after dying in Tarawa, Kiribati.

The details

The initiative mines databases across 14 languages to find evidence of service members missing since 1951. Investigators utilize these digital tools to navigate complex bureaucratic hurdles, including instances where foreign entities have charged as much as $10,000 for access to a single document.

1951: Earliest date of federal classified POW/MIA documents.
1990s: Senate Select Committee released 1 million declassified documents.
2015: Remains of Alexander Bonnyman Jr. were identified.
September 18, 2026: POW Investigative Project announced AI use.

The Tech Race

The use of AI in historical research marks a significant departure from manual document analysis, which previously limited the search scope to human reading speeds. This transition positions data science as a critical tool for solving historical mysteries long guarded by classified government repositories.

For the families of the missing, this technological push could provide the closure of identifying remains that were previously unreachable due to data volume. The project's efficiency gains may reduce the time and expense required for investigations into long-standing cases.

The takeaway

The deployment of AI for archival research demonstrates that modern processing power can unlock evidence trapped in massive datasets. This strategy offers a scalable model for government agencies and private researchers attempting to resolve decades-old disappearances.