The Department of Homeland Security (DHS) has recently introduced DHSChat, an AI-powered chatbot designed for internal use, marking a significant advancement in leveraging artificial intelligence to enhance productivity and support critical missions.
DHSChat is now accessible to approximately 19,000 employees at DHS Headquarters and select pilot users across ten operating agencies. This initiative builds upon the department's prior efforts, which permitted employees to utilize commercial generative AI tools like ChatGPT and Claude for publicly available information. Thousands of employees have received training and are currently using these tools to support their work. DHSChat utilizes similar AI technology but was developed internally to operate within a secure environment, enabling employees to perform routine tasks more efficiently, such as summarizing complex documents, generating computer code, and streamlining repetitive tasks like data entry. This tool allows thousands of employees to leverage generative AI capabilities safely and securely using non-public data.
The development of DHSChat is part of a broader effort by the DHS AI Corps, a team dedicated to recruiting seasoned AI experts and technologists to support AI development and deployment across the DHS mission. As of mid-December, thirty-nine of the fifty targeted experts have commenced work at DHS, contributing to over twenty AI-related projects since the AI Corps' inception in the spring.
DHS's commitment to responsible AI use is further demonstrated through the establishment of guiding principles outlined in Policy Statement 139-06, which ensures that AI systems, programs, and activities conform to the requirements of Executive Order 13960, promoting the use of trustworthy AI in the federal government. Additionally, Policy Statement 139-07 provides specific requirements and use limitations designed to safeguard data and protect privacy and individual rights when using commercial generative AI tools.
The department has also developed comprehensive training programs and interim rules of behavior to ensure that personnel are well-equipped to use generative AI tools responsibly. These resources emphasize the importance of meeting users where they are, discussing both expected and unexpected results, and providing access to approved tools to prepare the organization for the future.
DHS's proactive approach to integrating AI technologies reflects a commitment to innovation, quality, and value, aligning with FedSync's guiding principles. By developing and deploying tools like DHSChat, DHS is setting a precedent for the responsible and effective use of AI in federal agencies, enhancing operational efficiency while safeguarding privacy, civil rights, and civil liberties.