GAO Leverages AI Internally to Increase Efficiency and Effectiveness

The Government Accountability Office ("GAO") detailed "Planned Artificial Intelligence Use Cases" for leveraging new technology across federal agencies.

In testimony submitted to the Committee on House Administration, the GAO's Chief Data Scientist reported on the development of artificial intelligence ("AI") use cases "to make our work for Congress and taxpayers more productive, in-depth, and effective." The use cases include:

• organizing large volumes of material;
• summarizing documents;
• categorizing data;
• automated copyediting to focus on GAO's narrative structure improving consistency and productivity;
• writing automated chat responses to chat questions;
• survey analytics for quicker trend identification and interpretation;
• monitoring third party intent; and
• prioritizing work load.

The Chief Data Scientist said that one prototype the GAO is developing is a large language model "to explore generative AI capabilities like those found in industry, augmented with GAO-specific information and appropriate security controls." He said that the GAO has already analyzed more than 50 AI products, with 20 more AI projects on its agenda. In addition, he said the GAO is working on specific types of AI projects to include (i) cybersecurity and privacy of facial recognition technology, (ii) generative AI digital modeling, (iii) research for Department of Defense usage, (iv) federal workforce management, (v) national security and (vi) law enforcement efforts.