U.S. Intel Agency Takes Charge Of Mysterious “AI” Project

(FiveNation.com)- The National Geospatial-Intelligence Agency will take operational management of a critical Pentagon AI program, its head revealed.

As the Pentagon’s most public AI technology, Project Maven analyzes drone photos and video to identify prospective targets.

Under President Biden’s proposed budget for the Fiscal Year 2023, NGA Director Vice Adm. Robert Sharp indicated that his agency would transfer operational management of Project Maven’s GEOINT AI services from the Office of the Under Secretary of Defense Intelligence and Security.

According to Sharp, a platform and test and evaluation capabilities are included in this obligation.

The NGA is best known for processing and interpreting satellite data, but it also leads efforts to map the Earth and analyze other forms of aerial information.

Since 2017, the CIA has been closely involved in Project Maven.

At the GEOINT Symposium, NGA Deputy Director of Data and Digital Innovation Mark Munsell told reporters that started with this concept that commercial business – Silicon Valley – had come up with approaches dating back to 2008 or 2009. The aim was to bring in industry and commercialize what you’re doing. They would provide some military scenarios to test skills.

The Pentagon has used Project Maven to demonstrate its capabilities, particularly at Project Convergence, an annual Army exercise in the Arizona desert that tests sophisticated warfare technologies. The Army fitted an MQ-1C Gray Eagle with a Project Maven Maven Smart System and Algorithmic Inference Platform. The technology was one of several tested in the desert.

As part of Project Maven, NGA has provided data to firms to help them create AI algorithms. Project Maven’s primary focus is on geospatial intelligence, he says.
Munsell explains that finding a GEOINT home for everything they’ve learned, developed, and learned from the enterprises made a lot of sense.

According to Munsell, the NGA will not halt work when it takes over Project Maven, but it will bring in its experts to evaluate it and better integrate it with its other AI and computer vision projects.

The NGA has long sought to use AI to manage the vast amount of data from government and commercial sources. Officials believe there is too much data for humans to handle and rely on AI to identify trends that require human intervention. This notion of human-machine teaming is critical to the agency’s future operations.

Sharp said they are already leveraging AI and machine-learning techniques to allow large-scale GEOINT analysis. This move combines two significant DoD AI and ML machine learning programs. And they’re delighted that their tight collaboration with OSD employees helped produce Maven.