Latent AI Clears All Gates in U.S. Army Operation Jailbreak, Extends FTS Validation to Counter-UAS and Drone Autonomy

Field Tactical Suite proven on Lattice C2, NATO coalition stack, and unfamiliar partner hardware in three independent field tests

PRINCETON, N.J. — July 16, 2026 — Latent AI, a leader in delivering mission-critical intelligence at the tactical edge, today announced that its Field Tactical Suite (FTS) has cleared all four qualification gates in the U.S. Army’s Operation Jailbreak, earning a place on the U.S. Army API Marketplace and a path to a future sprint.

In the same period, FTS was independently validated in two additional field exercises: a NATO counter-UAS event with the Swedish Navy, and autonomous detection and tracking on unfamiliar partner hardware at Joint Interagency Field Experimentation (JIFX) 26-3. The three exercises tested FTS across different dimensions of readiness: command-and-control (C2) integration with the Army’s ecosystem and qualification gates, counter-UAS (C-UAS) interoperability inside a multinational coalition stack, and hardware-portable drone autonomy on unfamiliar partner hardware, all under live operational conditions

FTS is Latent AI’s edge-native AI deployment and adaptation suite. It enables teams to push, swap, and adapt AI models in the field across any edge compute hardware, without cloud connectivity, a return to the lab, or on-site ML expertise.

U.S. Army Operation Jailbreak: Command-and-Control Interoperability

Operation Jailbreak, launched in May 2026 at Fort Carson, Colorado, is the Army’s flagship sprint for integrating emerging technologies into C2 environments, with Anduril’s Lattice for Air Defense serving as the primary landing zone in support of the Integrated Battle Command System-Maneuver (IBCS-M). Clearing its gates means a vendor has exposed open APIs and proven interoperability with the Army’s C2 ecosystem, the bar the Army uses to assess readiness for wider fielding.

FTS integrated directly with Lattice, retraining models on new target signatures in the field and swapping the updated model into active service on operator command, without a cloud round-trip or a redeployment cycle. Read the full field account →

NATO Exercise with the Swedish Navy: Coalition Counter-UAS

FTS was deployed as the detection and tracking layer in a NATO SOFCOM counter-UAS event hosted by the Swedish Navy, integrated with the Trakka TC-300 sensor system, SeaCross C2, and NVIDIA edge compute hardware (a multi-vendor hardware and software stack Latent AI had not previously worked with) assembled under real weather conditions with a 48-hour integration timeline.

FTS ran UAS detection and tracking models across this unfamiliar coalition stack, serving as the AI layer bridging sensors, compute, and C2 from different vendors. In C-UAS, the detection layer is foundational: nothing downstream — tracking, classification, cueing — functions if the system can’t first see the threat. The exercise validated that FTS can plug into an unfamiliar multinational stack and be operationally integrated on a mission timeline, not a lab schedule.

JIFX 26-3: Drone Autonomy on Partner Hardware

At JIFX 26-3, the Naval Postgraduate School’s quarterly field experimentation event in Monterey, California, Latent AI ran a live integration test with Ceptor, field-testing FTS autonomous detection and tracking on Ceptor’s hardware, including a 13MP camera sensor running on the AR2020 processor, configurations the system had not been trained against. The model performed across real flight paths, varied terrain, and changing light conditions, providing the uncontrolled variables that a lab environment can’t replicate. The exercise produced a substantial dataset of operationally representative footage and validated that FTS detection capability holds up outside the conditions under which it was built. See the JIFX-to-Sweden Field Notes account here for the full narrative.

Todd Woodrick, President, Ceptor Technologies Corporation, said: “Ceptor Technologies Corporation develops specialized sensor systems for defense and space applications — environments where hardware performance and integration reliability are non-negotiable. What we look for in an AI partner is the ability to perform on our hardware, under field conditions, without a lengthy integration cycle. Latent AI brought FTS to JIFX and ran detection and tracking on Ceptor Technologies hardware it hadn’t been trained against — and it held up. That’s the kind of field-validated interoperability that matters for the programs we support.”  

Jags Kandasamy, CEO and co-founder, Latent AI, said: “Clearing every gate in Operation Jailbreak is the validation that matters most right now. It’s the Army’s own test of whether your AI belongs inside its ecosystem. That ecosystem is built on a modular, open systems approach (MOSA), and our entire business case rests on the same principle: AI that runs on the hardware the mission requires, in the environment it demands, without locking a program into a single vendor. Jailbreak, Sweden, and JIFX prove the same thing in three different settings — open, portable AI is what holds up in the field, on hardware you didn’t pick, with partners you didn’t plan for. That’s what we built FTS to do.”

About Latent AI

Latent AI is the trusted edge AI company delivering mission-critical intelligence at the tactical edge. FTS, the company’s Field Tactical Suite, enables defense and industrial organizations to deploy, adapt, and sustain AI in denied and contested environments — interoperable across platforms, field-updatable without cloud dependency, and built for operators of every skill level. Trusted by the U.S. Department of Defense. For more information, visit latentai.com.