Integrating Data Across Classification Levels A Challenge For AI

Integrating data across classification levels is the biggest challenge for the effective employment of artificial intelligence (AI) for U.S. military forces in the field, according to a new paper for the Modern War Institute at West Point by Tufts University Prof. Richard Shultz and United States Special Operations Commander Army Gen. Richard Clarke.

“Every step of the [data handling] process required direct human interaction,” per the paper, Big Data at War: Special Operations Forces, Project Maven, and 21st Century Warfare , which delves into the history of Project Maven before and after its kickoff in the spring of 2017. “While the process has gained some increased automation over time, the lack of a dedicated cloud-based data management infrastructure capable of quickly cutting across classification levels is the greatest roadblock to advancing AI capabilities for the warfighter.”

“Years of US military operations had left exquisite data repositories” for AI and machine learning, but “the data was fragmented across many silos” so that a system was unable to access multiple data repositories, and “data classification—both in terms of archival organization and security compartmentalization—had become a monumental roadblock,” the paper said.

U.S. Special Operations Command (SOCOM) said that it is working on a dedicated cloud-based data management infrastructure for special operations forces (SOF) that integrates data across security levels. “Our goal is to deliver solutions across the enterprise that enable SOF to leverage data however, whenever and wherever needed to meet operational requirements,” Navy Lt. Cmdr. Tim Hawkins, a SOCOM spokesman, wrote in an email. “Ongoing efforts to develop our cloud-based infrastructure and cloud native applications will accelerate progress, particularly as it relates to facilitating more seamless data exchanges between security domains, which will enhance support to SOF operators.”

Project Maven has looked to develop an AI tool to process data from full-motion video (FMV) collected by unmanned aircraft and decrease the workload of intelligence analysts.

Google [GOOGL] was the prime contractor for Project Maven but dropped out in 2018 after receiving pushback from employees about the company’s tools being used for an AI drone imaging effort. California-based big data analytics company, Palantir Technologies, co-founded and chaired by billionaire venture capitalist Peter Thiel, has assumed Google’s role, Business Insider has reported.

On May 20, 2017, former Deputy Defense Secretary Robert Work assigned to the Algorithmic Warfare Cross-Functional Team (AWCFT) under the Undersecretary of Defense for Intelligence the task of the automation of Processing, Exploitation, and Dissemination (PED) of tactical and mid-altitude full-motion video from drones in support of operations to defeat ISIS insurgents.

“Teams from a naval special warfare group were the first out of the gate,” according to the paper by Shultz and Clarke. “They were not immediately impressed with the algorithms’ performance but said they could see their potential. Later, the tools were also sent to USSOCOM’s counterterrorism teams conducting operations forward in other parts of the warzone. They had analogous reactions. While the AI could place a boundary box around vehicles, buildings, and people, and display them on a map, the algorithms were rudimentary with many false detections. In this process, called geo-referencing, analysts hoped to add location information to otherwise untagged data, so they could watch and track it. But that was not possible in the startup phase. Indeed, accuracy of detections was only around 50 percent. Determining the difference between men, women, and children was challenging.”

The role of AWCFT and U.S. Special Operations Command (USSOCOM) in fulfilling the goals established by Project Maven in 2017 “is a story that is still unfolding,” the paper said. “But when that story is told, it will not only be about automating FMV-PED to empower SOF [Special Operations Forces] teams to more effectively degrade and debilitate al-Qaeda and ISIS. It will also be the story of whether Project Maven served as the springboard to prepare DoD as an institution for future wars—a transformation from a hardware-centric organization to one in which AI and ML software provides timely, relevant mission-oriented data to enable intelligence-driven decisions at speed and scale. When that happens, U.S. commanders will be able to gain decisive advantage over current and future enemies.”

The Air Force is planning to integrate Project Maven into a test of the service’s Advanced Battle Management System (ABMS) in the next several weeks (Defense Daily, Aug. 12).

The Air Force has requested $3.3 billion for ABMS over five years, including $302.3 million in fiscal 2021.

ABMS, which the Air Force describes as the air and space “military Internet of Things,” is part of Joint All-Domain Command-and Control (JADC2), an effort to build a cross-service digital architecture for multi-domain operations.

The upcoming ABMS test is to be a “mega event,” Air Force Acquisition Chief Will Roper said, featuring all six military services and assets from space, air, land and sea, including space launch.