The United States Space Force has awarded a contract worth $27 million to Slingshot Aerospace to enhance the training of its personnel, known as Guardians. This agreement enables the implementation of the company’s TALOS AI system to simulate adversarial actions during orbital warfare exercises. The aim is to create realistic training scenarios that adapt dynamically to the decisions made by trainees.
According to Slingshot, TALOS has been trained on a comprehensive library of real-world orbital observations. This capability allows the AI to respond in a way that mimics real-world behaviors, rather than adhering to a pre-set script. Tim Solms, CEO of Slingshot Aerospace, emphasized the system’s adaptability, noting that TALOS can incorporate new algorithms and tools as they become available.
Integration of Advanced Simulation Tools
In addition to TALOS, Slingshot will supply various software tools that will simulate friendly (“Blue”) forces and serve as virtual referees (“White”). The contract, which spans 18 months, was issued through the Space Force’s Commercial Solutions Opening (CSO), a procurement method favored by the Defense Innovation Unit. This new contract builds on a previous $25 million award from 2022, which allowed the Space Training and Readiness Command to assess TALOS’s capabilities.
“This latest CSO award is a system-of-systems integration program supporting the Space Force Operational Test and Training Infrastructure (OTTI),” stated Solms. The project aims to consolidate various training tools into a unified solution, enabling Guardians to train within realistic and classified scenarios.
The TALOS AI system is designed to function as the thinking adversary, or “Red Cell,” within this integrated training environment. Additional tools that simulate friendly forces will also be incorporated, ensuring a comprehensive training approach.
Leveraging Extensive Data for Realistic Training
Launched in July 2025, TALOS aims to replicate satellite behavior in orbit for training purposes and can adapt as the orbital environment changes. Slingshot claims that its AI algorithms are trained on a significant volume of real-world data, tracking approximately 95% of all payload-sized objects across various orbital regimes, from Low Earth Orbit (LEO) to extended Geostationary Orbit (xGEO).
Slingshot’s capabilities allow for continuous tracking, providing data that the company asserts is the largest collection of commercially available astrometric and photometric data. “The resulting dataset is further enhanced by derived products that include event detections and patterns of life,” Solms explained.
The collaboration between the Space Force and Slingshot Aerospace marks a significant advancement in military training, utilizing cutting-edge AI technology to prepare Guardians for the complexities of modern orbital warfare. As space continues to evolve as a critical domain, initiatives like this will be key in maintaining operational readiness and effectiveness.
