The RQ-170 Sentinel — the stealth drone the CIA was flying over Iran — crashed near Kashmar in December twenty-eleven. Iran captured it mostly intact, reverse-engineered it, and within five years they were flying their own copy. China reportedly got access to the technology through that same channel. One crashed drone, two adversaries upgraded. That's the nightmare scenario that hangs over every F-35 sale.
And that's the thread Daniel's pulling in the prompt he sent us. Here's what he wrote.
According to recent media reports, the U.S. has decided that Turkey does not meet requirements to purchase the F-35. I'd like to talk today about the interesting question of restricted defense exports, specifically the F-35 and its cutting-edge predecessor. In a separate episode, we talked about the history of how Israel began customizing previous planes extensively — was the idea to take airframes that are almost hollowed out, rebuilt from the inside out? This decision by Israel raised significant friction with the U.S. at the time and almost led to the authorization being revoked. But other countries like the UAE and Turkey have been denied access to the technology at all.
He goes on — Beyond the U.S. context, we see Israel exporting systems like the Iron Dome to other countries, and these defense exports are a major driver of Israel's economy. The question is this: It's not hard to imagine a case in which an adversary would purchase an advanced weapon system through a third party or shell company solely in order to reverse engineer it — whether to gain their own defense capacity or to leverage that information offensively. So presumably these defense restrictions represent a balancing act between the market incentive to sell good technology that's the product of expensive R and D and recoup some of that cost, but on the other hand, not jeopardize national security through selling to either rogue actors or those intent on reverse engineering it. How do countries like the U.S., Israel, and others actually attempt to strike that balance in practice, at least from what we know?
That question — how do you sell a weapon so advanced it could be used against you — is exactly what we're going to unpack. And the Turkey case is the perfect entry point, because it's not a political fight dressed up as a security concern. It's genuinely about radar frequencies.
Let's start there. Turkey gets expelled from the F-35 program in twenty-nineteen. The headline version is: they bought Russian missiles, America got mad. What's the actual technical concern?
The S-400 is a Russian surface-to-air missile system, and its radar operates across a wide frequency band. The worry isn't that the S-400 shoots down an F-35 — the worry is that every time an F-35 flies anywhere near an S-400 radar, that radar is collecting data. Over weeks and months, you can build up a picture of what the jet actually looks like across the electromagnetic spectrum. You're essentially mapping its stealth signature.
So it's not a one-time exposure. It's cumulative.
Right. And the F-35's stealth isn't a binary thing — it's not invisible or visible. It's designed to look like a bird or a weather anomaly on radar. But if you know the specific frequencies where it glints, where the seams in the coating create returns, you can tune your air defense network to look for exactly those signatures. Turkey operating the S-400 alongside the F-35 would mean Russian engineers — who maintain the S-400 — are effectively standing next to the most sensitive stealth data the U.S. has.
And Turkey was a Level Three partner. They'd put a hundred and seventy-five million dollars into the program, they were supposed to buy a hundred jets, they were building components for the supply chain. This wasn't some distant customer — they were inside the tent.
Nine hundred parts were being manufactured in Turkey for the F-35. The U.S. had to unwind that entire supply chain over about eighteen months. It cost something on the order of half a billion dollars to re-source those components elsewhere. But the calculation was clear: the risk of the S-400 characterizing the jet's radar cross-section outweighed the economic hit and the diplomatic damage.
The mechanism the U.S. used was CAATSA — the Countering America's Adversaries Through Sanctions Act. What did that actually do?
CAATSA mandates sanctions on any country that engages in a significant transaction with Russian defense or intelligence sectors. The U.S. sanctioned Turkey's defense procurement agency, the SSB. The practical effect: the SSB couldn't access U.S. financial systems, its executives couldn't get visas, and any U.S. company doing business with them was also at risk. It's a blunt instrument, but it sent the message that the S-400 wasn't going to be grandfathered in.
And Turkey's response was basically — we'll build our own.
Which they're doing. The Kaan fighter program, the Bayraktar drones — Turkey's domestic defense industry has exploded since the expulsion. But that's a knock-on effect we should come back to.
So that's the baseline. A partner country makes a purchase that creates an intelligence vulnerability, and the U.S. pulls the plug. But that's the easy case — the threat was unambiguous. The harder question is what happens before the sale, when you're deciding who's trustworthy in the first place.
And that's where the formal requirements come in. Every country that wants to buy the F-35 signs a Memorandum of Understanding. It's not a handshake deal — it's a thick document that specifies exactly how the jet must be stored, maintained, and operated. Secure hangars with U.S. oversight. Encrypted maintenance data that's monitored. Any modification requires U.S. approval. There's a clause that basically says you cannot allow the jet to be exposed to hostile intelligence collection systems.
Which the S-400 triggered.
But beyond the MOU, there's the entire export control apparatus. The Arms Export Control Act governs foreign military sales. The International Traffic in Arms Regulations — ITAR — controls the technical data. The Defense Security Cooperation Agency manages the sales, but the State Department's Directorate of Defense Trade Controls has veto power. For something like the F-35, the National Security Council gets involved. It's a multi-agency process where any one of several players can say no.
What are the actual criteria they're evaluating?
Human rights record — that's a statutory requirement under the Leahy Laws. Regional stability impact — will selling this jet to Country A destabilize Country B? Ability to secure the technology — do they have the physical security, the counterintelligence capacity, the vetting procedures for personnel? And end-use monitoring capacity — can the U.S. verify, over the entire life of the system, that it's being used as agreed and hasn't been transferred?
End-use monitoring is the one that sounds like a bureaucratic checkbox but is actually the whole game.
It's where the trust meets the verification. Every F-35 has a mission data file that logs all flights and maintenance actions. The U.S. can remotely audit those logs. There are Defense Attaché offices at U.S. embassies that conduct on-site inspections. The supply chain is tracked — every spare part has a serial number and a chain of custody. And the software is designed to... well, the term is "brick" — if unauthorized modifications are detected, the jet can be disabled.
Wait. The U.S. can remotely disable an F-35?
The capability exists. The specifics are classified, but the architecture supports it. The F-35's Autonomic Logistics Information System — ALIS, though they're transitioning to a newer system called ODIN — that's the maintenance and operations backbone. It's connected. The jet phones home.
So you're not really buying an F-35. You're buying a subscription to an F-35.
That's... not wrong. The most extreme enforcement mechanism isn't sanctions or diplomacy — it's cutting off spare parts and software updates. Without those, an F-35 fleet is grounded within months. The jet is designed to be impossible to truly own. It remains tethered to the United States no matter whose roundel is painted on the wing.
Which brings us to Israel. Because Israel is the one country that said — we need this jet to be ours, and we need to put our own systems inside it.
The F-35I Adir. It's the only F-35 variant in the world that the U.S. allowed to be modified. Israel installed its own electronic warfare suite — the Elisra system. Its own data links. Its own air-to-air missiles, the Python-5. And this required the U.S. to do something it had never done: open up the F-35's source code and sensor fusion algorithms to foreign engineers.
That's the part that's staggering. The sensor fusion on the F-35 is the crown jewels. It's what takes data from six different sensors and merges it into a single picture that the pilot sees on the helmet display. Letting someone else look at how that code works is...
Unprecedented. And it almost didn't happen. The negotiations ran from about twenty-sixteen to twenty-eighteen, and there were multiple points where the U.S. side threatened to walk. The specific friction came when Israel wanted to integrate electronic attack capabilities that could potentially interfere with the F-35's own stealth systems. The U.S. had to verify — line by line, system by system — that the Israeli modifications didn't create vulnerabilities.
So let's address Daniel's question about the hollowed-out airframes. That image — an empty shell that Israel rebuilt from the inside — is that what happened?
Not literally. Israel received complete jets. But they were in what the program calls a non-standard configuration. Certain U.S.-classified systems were omitted at the factory and replaced with Israeli equivalents. It's more like... imagine buying a house where the builder agrees to leave the kitchen and bathrooms unfinished because you're bringing your own contractor. The structure is there, the walls are up, but key subsystems are yours.
And the friction Daniel mentions — the authorization almost being revoked — that was real?
Very real. There were members of Congress who argued that giving Israel source code access set a precedent that would be impossible to deny to other partners. The counterargument — which won — was that Israel faces existential threats that no other F-35 operator faces, and that Israel's intelligence cooperation with the U.S. is deep enough that the trust was warranted. But it was a close thing.
It's also worth noting that Israel's customization wasn't just about pride or industrial policy. Israel's air force operates in a neighborhood where the threats evolve on a different timescale than the Pentagon's upgrade cycle. If Hezbollah gets a new radar system from Iran, Israel needs to be able to tweak its electronic warfare response in weeks, not years.
That's exactly the argument the Israeli side made. And the compromise was a special Israel-specific software build that allows for faster integration of Israeli systems while still maintaining a U.S. audit trail. The U.S. can see what Israel is doing with the jet — it just can't stop them from doing it without pulling the whole agreement.
So Israel is the exception that proves the rule. Now let's flip to the other side — the countries that were denied. The UAE deal is the case that really illuminates the balancing act.
The UAE wanted twenty-three billion dollars' worth of F-35s. This was part of the Abraham Accords normalization — the Trump administration approved it in twenty-twenty. And then the Biden administration froze it in twenty-twenty-one. The sticking point wasn't the UAE's military capability or its relationship with Israel. It was China.
Specifically, Chinese 5G.
Huawei equipment at UAE military bases. The U.S. position was: you cannot have Chinese telecommunications infrastructure anywhere near where the F-35 operates, because that infrastructure can be used to collect signals intelligence. The UAE said — that's our sovereign decision. The deal collapsed.
Twenty-three billion dollars, walked away from, over 5G towers.
Which tells you everything about how the U.S. weights security versus economics. The F-35 program was explicitly designed to be cost-shared — partner nations contribute development funding in exchange for access and industrial participation. The whole model depends on exports to bring down the per-unit cost. And yet, when the security risk is clear enough, the economic argument loses every time.
That's the U.S. side. But Daniel's prompt also asks about Israel's export dilemma. Israel sells Iron Dome, Trophy active protection, drones. How does Israel's system work?
Israel's Ministry of Defense has a directorate called SIBAT — the International Defense Cooperation Directorate. They approve all defense exports. The framework looks similar to the U.S. system: end-use certificates, monitoring, restrictions on retransfer. But Israel has a unique constraint that the U.S. doesn't face.
U.S.-origin components.
Many Israeli systems contain American parts or were co-developed with U.S. funding. Under the third-party transfer provisions of the original sale agreements, the U.S. can veto Israeli exports. So Israel can develop the Iron Dome, but if it wants to sell it to a third country, the U.S. gets a say.
Has that actually happened? The U.S. blocking an Israeli export?
There have been cases. The Trophy active protection system — the one that intercepts anti-tank missiles — was co-developed with U.S. funding. When Israel wanted to export it to certain countries, the U.S. raised objections. It creates this weird dynamic where Israel owns the intellectual property but doesn't fully control the export decision.
So both the U.S. and Israel operate under similar constraints, but from opposite positions. The U.S. is the technology originator trying to maintain control; Israel is the technology adapter trying to maintain sovereignty.
And both are wrestling with the same fundamental tension: the better your weapon, the more you want to sell it to allies to recoup costs and build interoperability. But the more advanced it is, the more catastrophic the leak.
Let's talk about what those leaks actually look like. Daniel raised the shell company scenario — an adversary buying a system through a third party specifically to reverse-engineer it. Has that happened?
The RQ-170 you mentioned at the top is the cleanest case. Iran captured it, reverse-engineered it, and now flies the Saegheh and Shahed one-seventy-one drones that are direct copies. But there are other examples. Russia has been caught using shell companies in Eastern Europe to acquire Western missile guidance systems. In the early two-thousands, China obtained Israeli Harpy drone technology — that one actually caused a major diplomatic crisis between Israel and the U.S., because the U.S. believed Israel had transferred the technology without authorization.
The Harpy case is instructive because it wasn't a theft — it was a sale that went wrong. Israel sold the drones to China in the nineteen-nineties, and then China sent them back for upgrades in two-thousand-four. The U.S. argued that the upgrade constituted a new transfer that violated the original agreement. Israel eventually canceled the deal under U.S. pressure.
And that case directly shaped how Israel handles exports now. The lesson was: once the hardware leaves your control, you have to assume the buyer might do something with it you didn't authorize. So the control mechanisms have to be baked in before the sale, not enforced after.
Which brings us to the Turkey post-expulsion risk. Turkey spent years inside the F-35 program. Their engineers saw maintenance data, supply chain details, manufacturing processes. What's the actual risk that they could aid Russian or Chinese reverse-engineering?
The risk is real but it's not about the jet itself. Turkey doesn't have an F-35 to physically examine. What they have is institutional knowledge — they know which suppliers make which components, which materials are used in which coatings, how the maintenance diagnostics work. That kind of knowledge can accelerate someone else's development program by years. Instead of a Chinese engineer guessing what material the radar-absorbent coating uses, Turkey can tell them.
It's the difference between trying to reverse-engineer a cake by tasting it versus having the recipe but not the ingredients.
This is why the U.S. was so aggressive about unwinding the Turkish supply chain. Every Turkish company that had touched F-35 production was cut out. The concern wasn't just about Turkey — it was about Turkey becoming a conduit.
So how do you detect this kind of leakage? What's the actual monitoring apparatus?
Multiple layers. The Defense Attaché system — U.S. military officers at embassies who have inspection rights. The supply chain tracking — every component from a classified system has a unique identifier and a documented chain of custody. The remote auditing of mission data files. And then there's the intelligence layer — signals intelligence, human intelligence — that's looking for evidence that a partner country is sharing information they shouldn't.
And when you catch someone?
The ladder of escalation starts with diplomatic demarches, moves through restrictions on future sales, and tops out at the nuclear option: cutting off spare parts and software support. For the F-35 specifically, the U.S. can effectively ground a foreign fleet without ever touching the jets. No software updates means the jet falls behind on threat libraries. No spare parts means the readiness rate drops until the fleet is unflyable.
This is the thing I keep coming back to. The F-35 isn't just a plane. It's a platform that requires continuous connection to its manufacturer. And that's not a bug — it's the feature that makes export controls enforceable.
The C2D2 program — Continuous Capability Development and Delivery — pushes software updates every six months. New threat libraries, new capabilities, new countermeasures. If you're cut off from C2D2, your F-35 is frozen in time while the threats evolve around it. Within two or three update cycles, you're flying an obsolescent jet that cost you a hundred million dollars per airframe.
So the export control isn't just about the initial sale decision. It's a relationship that lasts for the forty-year life of the aircraft. Every software update is a renewal of trust.
And every software update is also a lever. The U.S. doesn't have to threaten to take the jets back — it just has to say the next update might be delayed.
Let's step back to the bigger picture. Defense exports are a massive economic driver. The U.S. did two hundred and thirty-eight billion dollars in foreign military sales between twenty-seventeen and twenty-twenty-one. Israel hit twelve point five billion in defense exports in twenty-twenty-two. These aren't rounding errors — they're major components of both economies. And yet we've just described a system where the U.S. walked away from a twenty-three-billion-dollar deal with the UAE over 5G towers. How do you make that calculus?
The calculus only makes sense if you think of the F-35 not as a product but as a strategic asset. The lifetime cost of the program is one point seven trillion dollars. The per-unit cost has come down — it's around eighty million now for the A variant — but the real value isn't the airframe. It's the stealth characteristics, the sensor fusion, the electronic warfare capabilities. If those are compromised, the entire U.S. tactical aviation fleet — not just the exported jets — loses its edge.
The math is: losing twenty-three billion in sales is cheaper than losing the deterrent value of a one-point-seven-trillion-dollar program.
That's the logic. Whether it's the right call in every case is debatable — the UAE deal had people arguing both sides — but the framework is consistent. The security tail wags the economic dog.
This creates a perverse incentive for countries that get locked out. Turkey is the clearest example — they're now investing heavily in domestic systems specifically because they can't trust the Western supply chain.
The Bayraktar TB2 drone has been a massive export success for Turkey. The Kaan fifth-generation fighter is in development. Turkey's defense industry is growing precisely because the F-35 door closed. S. technology may accelerate the development of competing systems that the U.S. has no control over at all.
It's the Streisand effect for fighter jets. The harder you try to keep someone out, the more determined they are to build their own.
When they build their own, they can sell it to whoever they want. Turkey has sold Bayraktars to countries that the U.S. wouldn't approve for F-35 sales. The export control regime works for the systems it covers, but it creates a market for alternatives.
Which brings us to the question Daniel's really getting at. How do you actually strike the balance? What's the framework that lets you sell enough to recoup R and D and build alliances, without selling so much that you arm your future adversaries?
I think there are three principles that emerge from the cases we've discussed. First, tiered access. Not every ally gets the same jet. The F-35 program has partner nations, Foreign Military Sales customers, and everyone else. The deeper your partnership, the more technology you get. Second, embedded control. The jet is designed so that even after sale, the U.S. retains veto power through software updates and spare parts. Third, continuous verification. The monitoring doesn't stop at the sale — it's perpetual.
The fourth principle, which is less formal but maybe the most important, is the trust network. The F-35 program is essentially a club. Members are vetted not just for their security practices but for their alignment with U.S. strategic interests. The Five Eyes countries — the deepest intelligence-sharing partners — get the deepest access. Israel gets unique access because of the depth of the intelligence relationship. Turkey got expelled because it broke the club rules.
That club model is fragile. It depends on every member believing that the rules are applied consistently and that the benefits of membership outweigh the sovereignty costs. If too many countries conclude that the U.S. will use its control levers arbitrarily, the club shrinks.
The UAE deal being frozen over Huawei — from the UAE's perspective, that looked like the U.S. moving the goalposts. The original deal didn't have a no-Huawei clause. It was added during the review.
That's the tension. The U.S. position is that the security environment evolves, and the conditions have to evolve with it. The buyer's position is that they made a deal and the seller is changing the terms. There's no clean resolution to that — it's an inherent feature of selling a weapon system that's continuously updated.
What's the next frontier? Daniel's prompt mentioned the F-35's cutting-edge predecessor, but I think the more interesting question is what comes after the F-35.
Software-defined weapons. The F-35 is already more software than hardware in terms of its combat capability. The next generation — the NGAD program, the collaborative combat aircraft — will be even more software-dependent. And that means the export control question shifts from "who can buy the jet" to "who can access the code."
You could imagine a future where the U.S. doesn't sell the weapon at all — it licenses access to it. The hardware sits in a U.S.-controlled facility, and the ally pays for the capability on a subscription basis. Revocable at any time.
That's already happening in some domains. Satellite imagery, signals intelligence — there are capabilities that the U.S. provides to allies as a service rather than a sale. Extending that model to combat aircraft is a big leap, but the technical foundation is being laid.
The F-35's real innovation might not be the stealth or the sensors. It might be that it's the first weapon system designed from the ground up to be impossible to truly own. It's a flying subscription service with afterburners.
That's either the future of defense cooperation or the thing that breaks the alliance system. I'm not sure which.
All right, let's go on the record. Predictions — concrete, checkable, with a timeframe.
Within three years, at least one major F-35 operator will publicly push back on a U.S. software update restriction, and it'll become a diplomatic incident that forces a formal negotiation over the terms of software access. The "subscription" model is going to be tested.
I'll go narrower. By twenty-twenty-eight, Turkey will fly a domestically-produced fifth-generation fighter — the Kaan — and will immediately offer it for export to countries that are locked out of the F-35 program. The export control regime will have created its own competitor.
That's a testable timeline.
The whole history of arms control is one step ahead of proliferation. The F-35's control mechanisms are impressive, but they're also a bet that the U.S. will remain the indispensable security partner for enough countries to keep the club viable. If that bet fails, the mechanisms don't matter — you've got a very expensive jet and nobody to sell it to.
The other side of that bet is whether the control mechanisms themselves drive potential partners away. Every time the U.S. pulls a deal over 5G towers or sanctions a partner's procurement agency, it's sending a signal: your jets work at our pleasure. Some countries will accept that. Some won't.
If you enjoyed this episode, please leave a review on Apple Podcasts or Spotify. It helps other weirdos find the show.
Thanks to our producer Hilbert Flumingtop.
This has been My Weird Prompts. We'll be back soon.
