Most robotics companies are dreaming big, but small. They want bots to fold your laundry. Or maybe stack shelves in a warehouse.
Sankaet Pathak disagrees.
He leads Foundation Future Industries, and he isn’t building household helpers. He wants an all-American supersoldier. A humanoid killer.
“We have some kinetic things we’re exploring” he tells WIRED, meaning weapons. “We’ll probably unveil something in the near future.”
He declines to give specifics, obviously.
Besides killing things, Pathak says the robots could handle logistics, recon, or inspections.
The Market is Bloody
The Pentagon likes this stuff. Always has. DARPA funded massive humanoid competitions from 2012 to 2015. The Army has xTechHumanoids, a program explicitly for “militarized humanoid capabilities.”
World militaries are rushing to adopt autonomous systems. Drones. Boats. Robots on legs. Legs can walk over trash piles and broken stairs better than wheels.
“If you look at Fallujah… you had several thousand insurgents hiding [in] buildings and troops just going door to door.”
A roboticist, speaking off-record, notes this makes tactical sense. If a robot takes the bullet meant for a marine, it’s a win. He’s surprised they aren’t in use yet.
Foundation says it has already tested its Phantom MK1 robot with Ukrainian forces. War, they argue, is a lab.
Then there’s the money. And the brand.
Eric Trump, president Donald Trump’s son, invests in Foundation Future Industries. He serves as chief strategy adviser, too.
“People don’t realize he actually is… an engineer,” Pathak says of his boss’s dad’s son. “He does a lot of milling at his home.”
On Fox Business in April, Eric Trump went off.
“…they fist-bump you. High-five. Follow commands,” he bragged. “The uses are unlimited. I think it’s very beautiful.”
Beautiful to who?
Foundation was born in 2024, barely a year ago. Then it bought Boardwalk Robotics, a group that worked closely with Florida’s Institute for Human and Machine Cognition, IHMC.
During that same Fox interview, the host claimed Foundation landed a “$24 million” Pentagon contract.
That sounds big. But the details get murky. When asked for proof, the company handed over records for contracts inherited from Boardwalk and others channeled through IHMC. Foundation itself appears to have zero direct new government contracts.
So the cash isn’t fresh. Yet.
Reality Check
Other experts roll their eyes. Fully autonomous soldiers are science fiction right now. At best, they are a distant maybe.
“Right now, it is very hard to tell what robots can actually do from what they could theoretically do,” says Robert Griffin of IHMC, a technical advisor to the startup. He notes that building an actual fighting robot faces challenges across every part of engineering.
Yes, motors are cheaper now. Sensors are better. AI can make these machines do parkour or kung-fu kicks.
But navigation? A disaster zone confuses them. Balance helps.
Then comes the physics of fighting.
Picking up a rifle, handling ammunition, reloading. Physical manipulation is a huge unsolved problem for robotics right now.
Rodney Brooks at MIT predicts it’ll take over a decade before these machines operate reliably anywhere but a controlled room. A warzone isn’t a controlled room.
“A lab demo turns into real-world use after about ten years,” says Brooks. Maybe longer.
“I can contribute to making war more precise. Reducing collateral damage.”
Pathak shrugs off the concerns about ethics and Terminator-style nightmares.
“Doomsday scenarios… very, very overblown.”
He thinks war will just become efficient. Precise. Less messy.
For the upcoming Phantom MK2, the upgrade isn’t better guns.
It’s waterproofing.
That’s where the money is going right now.
