The U.S. Space Force has awarded Slingshot Aerospace a $69.2 million contract to build artificial-intelligence training environments where Guardians can rehearse space-warfare missions against an autonomous AI opponent. Announced July 15, 2026, the deal is the largest in the Windsor, Colorado, company's history and runs under Space Systems Command's Operational Test and Training Infrastructure (OTTI) program. The 4.5-year award is a Small Business Innovation Research (SBIR) Phase III contract, the vehicle reserved for scaling technology that first matured under earlier SBIR-funded work.
Background
Slingshot Aerospace built its business around space-domain awareness, tracking objects in orbit and modeling their behavior. Over the past year the company has moved that capability into the training arena. An earlier $27 million Strategic Funding Increase (STRATFI) agreement supported development of the company's Digital Space Twin, a simulation of the space operating environment. The new $69.2 million award builds on that foundation and scales it into an operational rehearsal system for Guardians.
SBIR Phase III is a specific acquisition pathway. It lets an agency award follow-on production or services contracts, without renewed competition, to a small business that has already completed Phase I and Phase II research on the same technology. That structure rewards companies that carry an idea from concept through demonstration, and it explains how a firm outside the traditional prime-contractor tier landed a nine-figure Space Force deal. The Space Force has leaned on SBIR authorities repeatedly to bring commercial space startups into its orbit, and the Slingshot award is one of the larger examples of that approach paying out.
The program carries the name MENTAT, a reference to the human strategists in Frank Herbert's novel Dune who perform complex analysis without machines. The naming is deliberate: the system is meant to sharpen human decision-making rather than replace it. Guardians remain the decision-makers; the AI supplies the pressure, the variety and the pace that a live adversary would.
Key Details
At the center of the award is TALOS, short for Thinking Agent for Logical Operations and Strategy. Slingshot describes TALOS as an autonomous AI "virtual opponent" that models spacecraft behaviors and generates strategic response options without following a predetermined script. Rather than replaying canned exercise scenarios, the agent reacts to a Guardian's moves and produces new orbital situations on the fly, forcing operators to adapt the way they would against a live adversary. That is a meaningful departure from scripted simulators, which run the same way every time and stop teaching once a crew memorizes the sequence.
TALOS runs on Slingshot's Digital Space Twin, which fuses live orbital data, astrodynamics, AI analytics and mission context into a simulated operating picture. Because the environment is software-based, crews can rehearse protect-and-defend missions from their home station instead of waiting for scheduled live exercise windows, which are limited and expensive to run.
The contract runs 4.5 years under the OTTI program office at Space Systems Command, the Space Force organization responsible for the service's test and training infrastructure. OTTI's mandate is to give Guardians realistic, repeatable environments in which to practice, and the Slingshot award slots directly into that mission.
"Talos gives Guardians the ability to train against realistic orbital scenarios and rehearse complex missions with a level of speed, scale, and accessibility that simply hasn't existed before," Slingshot CEO Tim Solms said in announcing the award. The company frames the technology as a shift from occasional, resource-heavy exercises toward continuous, on-demand training that a crew can run whenever it needs the repetitions.
The 4.5-year contract lands as the Space Force sharpens its focus on orbital warfare, a mission the service increasingly frames around protecting its own satellites and countering hostile maneuvers by rival spacecraft. Training crews to recognize and respond to those threats has become a priority, and until now much of that rehearsal depended on scarce live exercises. The Slingshot system is meant to close that gap by putting a thinking opponent in front of Guardians on demand.
What It Means for Contractors
The award is a case study in how SBIR Phase III can carry a small business into large-dollar defense work. Slingshot did not win this contract through a full-and-open competition against primes; it positioned the capability with earlier SBIR research and a STRATFI funding agreement first, then used the Phase III authority to scale. For small and mid-size firms, the lesson is that early SBIR and STRATFI positioning on a specific technology can convert into a sole-source follow-on years later, provided the underlying work traces cleanly back to the original research.
The award also signals where Space Force training dollars are heading. The service is investing in autonomous, AI-driven synthetic opponents and software-based rehearsal rather than only physical ranges and scripted exercises. Vendors with strength in astrodynamics, orbital modeling, AI analytics and simulation now have a clearer demand signal from OTTI and Space Systems Command. Companies holding relevant space-domain data or modeling tools should watch the OTTI program office as a customer, and firms that supply pieces of a simulation stack — data feeds, analytics, scenario generation — may find subcontracting room as Slingshot builds out MENTAT over the coming years.
Slingshot has said the award will let it expand the fidelity and scope of its mission-rehearsal environment, layering more realistic adversary behavior on top of the orbital data it already collects. The company positions the Digital Space Twin as a foundation it can keep building on, meaning the MENTAT work could seed still more follow-on business if the service is satisfied with the results.
Finally, the deal reinforces that the Space Force is willing to place nine-figure bets on non-traditional vendors when the technology fits. That is an opening for the broader small-business space sector, but it comes with a caveat: the winners are the companies that invested early, matured the technology through the SBIR pipeline, and kept the paperwork trail intact so a Phase III award could stand on its own. For contractors weighing whether to pursue Space Force training and simulation work, the message is that a credible orbital-modeling capability, backed by a documented SBIR history, is now a path to serious money.
