Artificial Intelligence and Machine Learning

Through both deep institutional knowledge and high-quality, intersectional research capabilities, RAND is uniquely equipped to help the Department of Homeland Security and the Homeland Security Enterprise meet the threats and assess the opportunities posed by AI and machine learning to keep the American homeland safe.

The Challenge

The explosive growth of AI tools—machine learning systems, generative large language models, and, someday, general artificial intelligence—expose a new frontier of threats and opportunities for the United States. Emerging technologies will have effects across every Department of Homeland Security mission, from countering terrorism to securing cyberspace to preserving American economic security.

Photo by LeoPatrizi/Getty Images

RAND's History

For nearly eight decades, RAND has played a pivotal role in advising government agencies about new technologies: in 1946, it released its first report on the potential design, performance, and use of manmade satellites; in the 1970s, RAND explored the previously unthinkable—unmanned aerial vehicles; and in the 1980s, RAND designed the building blocks for today's internet.

Mission Support

Artificial Intelligence (AI) and Machine Learning (ML) can increase the speed, accuracy, and convenience of many of the Department of Homeland Security's core missions. But leaders are rightfully cautious of the potential risks associated with the use of emerging technologies—discriminatory results, a lack of transparency and oversight, and the infringement of privacy and civil liberties.

Researchers from RAND have been exploring the use of emerging technologies by DHS and other government agencies for the past several years. A suite of ongoing research projects are examining future uses of AI and ML, along with public perceptions about emerging technologies. Other core RAND capabilities in AI/ML include:

  • Applying machine learning for prediction, classification, and anomaly detection
  • Tailoring large language models for security applications
  • Evaluating machine learning systems across the lifecycle
  • Developing policies related to AI deployment, privacy, and risks
  • Educating policymakers about AI, machine learning methods, and their implications

    • On this page, you'll find a sampling of recent work on AI and ML curated specifically for the homeland security community. To view RAND's full archive of work on AI and ML, visit RAND's Artificial Intelligence Topic Page.

    Featured Research

    For more projects, see our topic pages on Artificial Intelligence and Machine Learning.