The Same Gear. Two Very Different Contexts.

While a New York Post reviewer was logging 30,000 steps around Prospect Park with a consumer-grade AI exoskeleton strapped to her waist, Ukrainian soldiers were using nearly identical-looking hardware to load artillery shells in the Pokrovsk region under active fire.

This is the current state of exoskeleton technology in 2026.

What's Happening on the Ground in Ukraine

According to BGR, Ukraine's 7th Air Assault Corps posted a Facebook video in April 2026 showing soldiers strapping on AI-enabled mechanical waist and leg braces before executing combat tasks — including loading a CAESAR self-propelled Howitzer.

The stated purpose: reduce physical load on soldiers while boosting speed and stamina. The unit confirmed the technology is being tested in both logistical and combat operations.

Ukraine is using a version of the exoskeleton that consumers can buy on Amazon right now, according to BGR.

The U.S. Army Isn't Sitting This One Out

According to the Daily Mail, the U.S. Army is actively testing AI exoskeletons on soldiers. The report offers few specifics, but it confirms American military interest is real and ongoing.

This isn't new territory. Lockheed Martin secured a development agreement for its ONYX exoskeleton nearly a decade ago, per BGR. The U.S., China, and Russia have all pursued the technology. What's changed is the pace.

The war in Ukraine has compressed what would have been a decade-long R&D cycle into a live-fire testing sprint. Depending on your confidence in the procurement process, that's either promising or concerning.

The Science Behind the Suit

Mainstream coverage largely skips why this technology didn't exist at scale five years ago, and why it's viable now.

According to the U.S. National Science Foundation, the biggest obstacle to widespread exoskeleton use has always been individualized calibration — the process of programming a suit to work with one specific human body. Current methods take 30 to 60 minutes per user in a lab setting. That's a non-starter for battlefield deployment.

Researchers at North Carolina State University, funded by NSF, cracked the problem. They developed an AI-driven simulation framework — a learning-in-simulation model incorporating human, robot, and human-robot interaction data — that eliminates the need for each user to go through lengthy individual calibration. NSF program director Alexander Leonessa called it "a major advance in exoskeleton engineering."

One control policy. Multiple users. Multiple activities — walking, running, stair climbing. No lab required.

This breakthrough is what makes the jump from expensive prototype to deployable technology possible. Without it, you're still fitting every soldier for a bespoke mechanical suit. With it, you can issue exoskeletons like body armor.

Meanwhile, the Pentagon Is Testing AI Logistics in the Baja Desert

The exoskeleton story doesn't exist in isolation. There's a broader push to field AI-assisted physical performance tools for military use.

According to Defense One, GDIT is partnering with AWS on a project called Project Celerity — an AI platform for managing energy in disconnected, contested environments. They're field-testing it in the upcoming Baja 1000 dirt-bike race, which GDIT's VP for AI and machine learning, Brandon Bean, called "a proxy for contested logistics."

The Army's Advanced Research Lab is simultaneously investing in tactical microgrids — portable power systems for remote operations — to keep a growing fleet of electric and robotic vehicles running. Predicting battery needs for drones and ground robots in real time is part of the challenge, per Defense One.

AWS's Shannon Judd confirmed the military applications include special operations forces, remote patrol units, and disaster response missions.

Power management. Predictive maintenance. Autonomous logistics. These aren't science fiction concepts anymore. They're being tested in dirt and sand this year.

What the Coverage Is Getting Wrong

Most mainstream outlets are treating these stories as separate: the cool consumer gadget review, the Ukraine battlefield curiosity, the Pentagon logistics experiment, the NSF research grant.

They're connected. The NSF-funded calibration breakthrough makes mass deployment feasible. Ukraine's battlefield testing provides real-world performance data at zero cost to U.S. taxpayers. The consumer market funds continued R&D. The GDIT/AWS energy AI solves the power problem that limits how long an exoskeleton soldier can operate in the field.

Every piece is feeding the others. The Daily Mail's coverage of U.S. Army testing is frustratingly vague — no unit names, no program names, no dollar figures. The NY Post review, while honest and readable, frames this entirely as a fitness product with no acknowledgment that the technology has direct military applications.

What This Means for Regular People

In the short term: AI exoskeletons are real, they work, and they're already available to civilians for fitness use. The Hypershell X Ultra S — which the NY Post reviewer tested — uses 12 sensors and a real-time motion prediction algorithm. It's not a gimmick.

In the medium term: the same technology is going to show up on battlefields at scale. American soldiers, Ukrainian soldiers, and almost certainly Chinese soldiers will be wearing AI-assisted exoskeletons within this decade. The calibration problem is solved. The power problem is being solved.

In the long term: the line between consumer tech, military tech, and medical assistive devices is erasing. North Carolina State's NSF-funded framework was designed to help both able-bodied users and mobility-impaired individuals. The same breakthrough that lets a soldier carry artillery shells longer could let a stroke patient walk again.

The technology doesn't distinguish between applications. This convergence is why everyone — from the Pentagon to Amazon shoppers — should be paying closer attention.