NGA’s $700 million data labeling RFP to enhance AI for GEOINT
National Geospatial-Intelligence Agency is preparing a $700 million data labeling contract aimed at advancing AI-driven computer vision models for identification of military targets. This marks the largest data labeling RFP in U.S. government history, according to GovConWire. Key to NGA’s initiative is Project Maven, the agency's flagship AI program, to process an ever-growing amount of geospatial intelligence (GEOINT) data from satellites.
As NGA Director VADM Frank Whitworth noted, the influx of data presents computational challenges that can slow AI processing capabilities. Data labeling, the process of manually identifying and distinguishing objects in images for AI training, is critical for improving the models’ accuracy, especially in high-stakes scenarios like military targeting.
Companies interested in bidding should monitor this RFP and attend the 2024 Intel Summit, where key decision-makers will provide additional insights into the agency’s AI and data labeling strategies.