Foundation Models at the Department of Homeland Security: Use Cases and Considerations

Considered the modern-day backbone of artificial intelligence (AI), a foundation model (FM) is a type of machine learning model that is trained on a broad set of general domain data for the purpose of using that model as an architecture on which to build multiple specialized AI applications. By collapsing data and technology across use cases, FMs benefit from increases in the scale and scope of datasets to become more capable and from economies of scale in workflow to become more efficient and impactful. To better understand how we can learn from these FMs with respect to images and other data sources, which inform decision-making, this analysis examines a host of different homeland security use cases, paving the way for a safer and more resilient nation.