Humans and Machines, A Match Made In and For the Heavens: How the U.S. Space Force Leverages AI

By: Samantha Louque, Marketing and Outreach Project Coordinator at the Innovation Hub

“The Space Force needs to emphasize AI literacy to employ these (AI) tools to solve operational challenges.” That’s according to the Space Force’s Data & Artificial Intelligence FY 2025 Strategic Action Plan, which outlines four lines of effort aimed at advancing AI culture, adopting new AI tools, and promoting collaborative and innovative thinking. 

However, many of our Space Force guardians have concerns about letting machines control aspects of our national defense in space. Artificial intelligence tools are rarely developed with guardians in mind. Plus, it doesn’t take a Terminator fan to recognize we shouldn’t let machines take over our weapons systems. 

Fortunately, some of the smartest minds across industry, academia, and government are working towards optimizing human-machine teaming in the Space Force. By enhancing AI systems to partner with our guardians, not replace them, the U.S. can move towards a more efficient and secure space domain. 

Modern AI systems began to soar in popularity in 2023. Suddenly, we could speak to computers in our language without having a background in computer science or coding. The barrier to entry for once-complex computer systems became much lower. Now, AI tools can help with everything from drafting monotonous emails to processing massive datasets. 

Unfortunately, U.S. adversaries have also benefited from this technological revolution. China is significantly increasing its defense spending, investing in science and technologies such as AI tools. The nation has already begun launching satellites for their planned AI supercomputer constellation in space. 

As our adversaries ramp up their use of AI tools, fears around the unethical use of this technology have increased. A New York Times article reported that “The Chinese government is using companies with expertise in artificial intelligence to monitor and manipulate public opinion, giving it a new weapon in information warfare.” American intelligence agencies are currently monitoring for this influence to expand into U.S. political spheres. For example, the Department of Justice has already disrupted AI-driven Russian disinformation campaigns. 

Recognizing the need to thoughtfully exploit emerging AI technologies and close gaps between the U.S. and our adversaries, the Space Force recently established the Information Dominance Tools, Applications, and Processing (TAP) Lab. The Innovation Hub (IHub), a collaboration facility in Colorado Springs established as a partnership between the Virginia Tech Applied Research Corporation (VT-ARC), Space Systems Command (SSC) and the Air Force Office of Scientific Research (AFOSR), hosts these various TAP Labs as well as 20 additional government teams. For the past couple years, SSC has utilized TAP Labs to rapidly develop capabilities necessary to meet mission needs. (See previous AG Coverage of the IHub). Now, the IHub also serves as the center of mass for the new Information Dominance Lab. It offers resources such as computer science support, event coordination, and business intelligence services that bring members of the Lab together to thrive.  

The goal of the Information Dominance TAP Lab is to follow data from the moment it’s ingested into the digital ecosystem, across the information exchange boundary where it’s manipulated and presented to humans, all the way to the point where humans internalize that information and implement the decision-making process. As that’s a long and complicated pathway, its founders recognized the need to take two different approaches to bring it all together. They broke down the Information Dominance TAP Lab into two sub elements: the AI TAP Lab and the Human-Machine Teaming TAP Lab. 

The AI TAP Lab

Before humans and machines can start effectively operating together, we must ensure that our AI systems are up to military standards. “Most AI tooling is not generated with specific military ends in mind,” noted Bart Stewart, S6 Command Data and AI Officer for SSC and a founder of the Information Dominance TAP Lab.

Thus, the AI TAP Lab’s first objective is to mature and adapt AI systems to fit the purpose of military applications. Ultimately, AI tools should have a specific purpose in battle management decision support. This starts with training systems on synthetic data, where users test AI models to see if they deliver the predicted results. Throughout that process, users will develop the underpinnings and infrastructure that allows the Space Force to take leading industry tech and apply it to military problems in the space domain, ranging from improving acquisition efficiency to enhancing operators’ ability to deliver lethal effects.

The more we refine and develop underlying systems using these techniques, the more humans will trust AI models to perform as expected. “Building this trust relationship is the same thing we do with human operators,” noted Stewart. “We send guardians to basic training to build trust in them to perform as expected. We need to be able to do that with AI models.” 

The Human-Machine Teaming TAP Lab

Once the Lab and its partners bring AI tools up to military standards, the Human-Machine Teaming TAP Lab will focus on closing operational gaps through automation. This will empower operators to offload mundane and repetitive tasks that take up time and energy. By lowering guardians’ cognitive load, AI leaves them more brain space for making important decisions. 

By thinking of AI as a member of the team, Space Force personnel can distribute tasks to optimize functionality. As with any good working relationship, it’s important to clearly define which partner is best at which function. Machines are particularly well equipped to handle tasks that require a large amount of memory, redundant or repetitive data entry, or postulating multiple potential future outcomes from an action. On the flip side, making decisions around lethal force that involve morality, rule of law, doctrines related to armed conflict, and the art of battle management should be left to humans.

The Information Dominance TAP Lab has already started exploring how data fusion tools can simplify decision making in the Space Force. The Space Domain Awareness (SDA) TAP Lab, another Innovation Hub government partner, has already implemented Data Fusion into their kill chain. (See previous AG coverage of the SDA Tap Lab). Data Fusion, the process of combining information to improve its quality, reduce uncertainty, extract important features, and provide statistical and analytical insights, has already helped the SDA TAP Lab guide project data and evaluate example scenarios. 

The Information Dominance TAP Lab wants to take this a step further using a Bayesian Committee Machine (BCM), a data fusion tool. BCMs are artificial intelligence tools designed to mathematically simulate a human committee in the decision-making process. Those who utilize BCMs primarily do so to train other AI systems, but the Innovation Hub’s data science team is researching novel use cases. Together, VT-ARC and the SDA and Information Dominance TAP Labs are working to determine whether BCMs can supplement decision making for some of the SDA TAP Lab’s processes. 

Like all innovative programs, the Information Dominance TAP Lab doesn’t operate in a silo. The Lab works closely with the U.S. Space Force Chief Scientist Dr. Stacy Williams, whose team focuses on researching future science and technology solutions that can ultimately be incorporated into the Lab. The Warfighter Effectiveness Research Center (WERC) in the U.S. Air Force Academy (USAFA) serves as the academic integrator for the Lab. It brings in new ideas and research. Finally, although the Lab is a Space Force program, it draws on resources from across other branches of the DoD. 

The Innovation Hub provides a setting where these teams from across industry, academia and government can come together to solve our nation’s most pressing AI challenges. Together, they’ll ensure that the United States stays at the forefront of AI and human-machine teaming and makes the space domain both smarter and safer.  

Effort sponsored in whole or in part by the Air Force Research Laboratory, USAF, under Partnership Intermediary Agreement No. FA9550-22-3-0001. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of the Air Force Research Laboratory.