Artificial intelligence doesn't wait. It improves in weeks, sometimes days. The Pentagon, on the other hand, has historically bought weapons on timelines measured in years, even decades. That mismatch is becoming one of the biggest headaches for the U.S. defense establishment, according to Aviv Shapira, CEO of Xtend AI Robotics, Inc. (NYSE: XTND).
Shapira isn't just some outside critic. His company builds AI-driven drone systems for defense, so he's got skin in the game. He gives the Pentagon credit for recognizing that AI and autonomy matter. But he says the real test is whether the acquisition system can keep up with tech that moves at startup speed.
"The Pentagon is in the frontline of AI and has made significant progress in adapting to the pace of AI and autonomy, and we're seeing that shift happen in real time," Shapira told MarketDash in an exclusive email interview.
The problem, he explains, is that the current procurement system was built for an era when a tank or a fighter jet stayed relevant for years with only minor tweaks. AI shatters that assumption. The technology changes fast, and so do the threats it's meant to counter.
"The challenge now is that AI and autonomy break that mold, as both the technology and the threats it counters can shift significantly within days and weeks instead of months and years, so acquisition must move at the same pace," Shapira said.
So what's the fix? Shapira says it's not just about speeding up the existing process. It's about changing the whole mindset. Instead of buying hardware and calling it done, the Pentagon needs to think of defense systems as software that gets better over time.
"The next step is making procurement more software-native, with shorter contracting timelines, faster testing and evaluation, stronger pathways from successful prototypes to scaled programs, and acquisition models that support continuous software upgrades," he said.
That's a big shift. For decades, defense acquisition has been about big, one-time purchases. You buy a platform, you use it for 30 years, maybe you upgrade the radar or the engine. But AI-enabled systems are different. They improve through software updates, the way your phone gets better with each new iOS. The hardware is almost secondary.
Shapira argues that the old model, built around periodic hardware upgrades, just can't capture the full value of AI-driven platforms. The real measure of success, he says, isn't how fast a system gets to the field. It's how fast it gets better once it's there.
"Ultimately, success is measured in how quickly a deployed unit can turn an operational lesson into a verified capability improvement in the field," he said. "That is the pace AI and autonomy can enable, and the Pentagon increasingly understands that."
This isn't just an academic debate. It has real implications for investors. If the Pentagon shifts toward software-native procurement, companies that can continuously update their AI platforms and operate within faster, more agile acquisition models could be the big winners. Defense spending is already moving toward autonomous and software-defined capabilities, and the companies that fit that mold might be better positioned than those stuck in the old hardware-centric world.
Shapira's comments suggest that the next phase of defense modernization might hinge as much on procurement reform as on technological breakthroughs. The tech is ready. The question is whether the bureaucracy can catch up.
