This article was originally published on Technology.org.
I returned to the U.S. about two weeks ago, after spending the better part of a year in Israel covering Israel’s war with Hamas in Gaza and now the confrontation with Iran. A year is a long time to be embedded in a country at war. You see things that don’t make the evening news. You develop a sense of how a society actually operates under sustained threat — not in theory, but day to day.
One thing struck me more than anything else: artificial intelligence in Israel is not a buzzword. It’s not a pilot program or a Silicon Valley talking point. It’s infrastructure. As common and as load-bearing as the country’s electrical grid.
When I came home, I started calling sources to make sense of what I’d witnessed. I spoke with Elliott Broidy — entrepreneur, CEO of Broidy Capital Holdings, and founder and managing partner of LEO Technologies, LLC (LeoTech), a firm deploying AI tools for law enforcement and public safety agencies across the United States, and Chairman and CEO of agentic AI Video Intelligence company Cynapse AI. Broidy has spent years watching the intersection of AI and national security from the investment side, and he has a clear-eyed view of where Israel sits relative to the rest of the world.
“What Israel has done,” Broidy told me, “is to compress decades of technological development into a few years of operational necessity. When your survival depends on making better decisions faster than your adversaries, you don’t wait for a perfect product. You build, you test, you deploy — and you refine under pressure.”
That compression is exactly what I observed on the ground.
A Hospital That Functions Two Stories Underground
One of the more surreal moments of my time in Israel came during a visit to the Sheba Medical Center outside Tel Aviv — the largest hospital in the Middle East, with 11,000 employees, 1,700 doctors, and over 3,000 nurses. What I found was a functioning hospital operating almost entirely underground, out of reach of Iranian ballistic missiles.
Pediatrics, oncology, ophthalmology, cardiology — all of it had migrated beneath the surface. Ambulances unloaded patients on gurneys in underground bays while workmen fixed overhead cables nearby. “This place is not a field hospital. This is the hospital. It is just subterranean,” one pediatric doctor told me. She said the staff had assumed they might be working this way indefinitely.
After the 12-day Israeli-U.S. war with Iran last June, hospital leadership had studied what they’d learned and built a more distributed underground system — spread across five subterranean locations, each capable of running independently.
What made it possible was the AI-driven systems, which monitor threat levels, coordinate logistics, and help hospital command know when to move, how fast, and in what sequence. The military gave the hospital roughly a month’s notice before the latest air strikes, based on its own predictive modeling. That lead time saved lives.
Predicting Emergencies Before They Happen
Below the level of military operations, Israel has also wired its civilian emergency response system with AI in ways that feel almost sci-fi.
Members of Unit 8200, Israel’s elite signals intelligence unit, collaborated with the emergency response organization United Hatzalah to build an AI dispatch system trained on 18 years of historical emergency data. The system predicts where medical emergencies are likely to occur within the next 10 minutes, dynamically repositioning ambulances and first-response “ambucycles” ahead of actual calls. During a three-month pilot, it correctly predicted the location of the next emergency call — within a one- to two-minute response radius — roughly 85 percent of the time.
The implications are straightforward: faster response times, more lives saved. In one documented case, a volunteer was directed to a mall based on an AI prediction. Before he arrived, his app flagged a patient with chest pains 400 meters away. He reached the patient before the caller even hung up the phone.
Magen David Adom, Israel’s national emergency service, has built a similarly sophisticated system — a Computer-Assisted Design AI platform that simultaneously alerts and dispatches the nearest certified civilian responders, ambulance crews, police, and fire services the moment an emergency call comes in. It routes everyone using real-time traffic data and predictive algorithms, all without a dispatcher manually coordinating each agency.
“What Israel has built in the civilian sector is just as instructive as the military applications,” Elliott Broidy said. “The same logic applies: you reduce the time between a threat and a response, you reduce harm. Whether the threat is a cardiac arrest or a missile, the principle is identical.”
On the Battlefield: AI as Operational Doctrine
The IDF’s adoption of AI goes well beyond logistics. Deep inside a fortified command facility known as the “Fortress of Zion” — a bunker beneath IDF headquarters in Tel Aviv — military planners coordinate air, sea, and land operations from a nerve center built for high-tech warfare.
On its walls of screens: three-dimensional building diagrams, live aerial drone feeds, real-time maps of ground and air assets across the region. During combat operations over Gaza and Lebanon, roughly 300 to 400 personnel worked there around the clock, with that number swelling to the thousands during active campaigns. Intelligence analysts from Mossad, Shin Bet, the Foreign Ministry, and police all shared the same operational picture.
The IDF has used this infrastructure to target enemy assets with a precision that older command structures simply could not achieve. It has also grappled with the serious ethical questions that come with it — the distance between a commander and a target, the reduction of human lives to data points in a system, the risk of indifference that can creep in when war is conducted from a climate-controlled room. Senior IDF officers acknowledge this tension directly. It is, as one general put it, “part of the commander’s challenge.”
On the medical side, the IDF Medical Corps has been running its own AI-driven revolution. An R&D division within the corps reduced evacuation times from the Gaza battlefield to a hospital to an average of 60 minutes — well within what medics call “the golden window” for saving the critically wounded. In Lebanon, where distances were greater, the average was 107 minutes. Paramedics were pre-positioned close to front lines so they could reach wounded soldiers within two to three minutes.
The corps also developed a drone delivery system for whole blood — one of the most time-sensitive medical supplies in combat. Engineers spent years solving the physics: keeping blood at two to four degrees Celsius at altitude, designing landing systems that wouldn’t rupture bags on impact, ensuring drones flying at 200 meters wouldn’t be shot down by friendly forces who didn’t recognize them. By October 2025, a full-scale drill had worked. Operational use was expected shortly after.
AI has also entered the realm of mental health. Using EEG machines paired with machine learning, IDF mental health professionals can now map individualized “calming cues” for soldiers dealing with PTSD, identifying whether exposure to a particular song, voice, or image has a measurable neurological effect. Separately, AI tools are cutting the time clinicians spend on administrative record-keeping from an hour down to about 15 minutes per session, freeing up more face time with patients.
What the U.S. Can Learn
Elliott Broidy has thought carefully about the gap between where Israel operates and where the United States still aspires to be.
“Israel doesn’t have the luxury of bureaucratic timelines,” he said. “When you’re a country of 10 million people surrounded by adversaries on multiple fronts, every tool has to work. That’s created a culture where AI gets deployed, gets tested in the hardest conditions imaginable, and gets refined fast. The U.S. has enormous assets, but we don’t always move with that kind of urgency.”
His firm, LEO Technologies, is trying to apply that urgency in the American public safety context. Through its VerusAI platform, LeoTech brings agentic, semantic, and generative AI to law enforcement agencies — compressing what used to take investigators weeks of manual work down to seconds. The platform monitors communications for contraband coordination, gang activity, and suicide risk in correctional facilities, surfacing the intelligence that matters and filtering out the noise. It currently serves more than 224 agencies nationwide.
“The core insight is the same whether you’re in Tel Aviv or Tulsa,” Broidy said. “AI doesn’t replace human judgment — it gives humans better information, faster, enabling better decisions that can be deployed more quickly. The question is always: what does your team need to know, and how quickly can you get it to them?”
A Country That Builds Under Fire
What I keep coming back to, weeks after returning home, is the sheer normality of AI’s role in Israeli life. Emergency dispatchers rely on predictive algorithms the way American dispatchers rely on phone maps. Hospitals have AI-integrated contingency protocols baked into their operations. The military runs targeting and logistics through systems that would have seemed like science fiction a decade ago.
The pressure that produced this ecosystem — decades of conflict, constant threat recalibration, a small population that cannot absorb large losses — is not something any country would choose. But the technological culture it has built is worth studying seriously. Israel has figured out something that larger, more comfortable nations are still working through: that AI deployed under real conditions, with real stakes, matures faster and proves its worth more clearly than AI developed in peacetime labs.
As Elliott Broidy put it: “Israel is the world’s most advanced proving ground for AI in national security. What works there, works. And what works there today often becomes doctrine everywhere else tomorrow.”