Daniel sent us this one — and it's timely. In March, Israeli intelligence was reported to be arming the Abu Shabaab clan in Gaza, small arms, night-vision equipment, the works. The idea being to build a counterweight to Hamas. And that raises the question at the heart of this prompt: when governments arm unofficial militias and insurgent groups, how do they actually do it, and how do they try to make sure those guns don't end up pointed back at them? And the second part — do governments always formally deny they're doing this? Because it sure seems like the answer is yes.
The Abu Shabaab case is a perfect entry point because it's happening right now, but the playbook is ancient. I mean, arming someone else's enemies is basically statecraft one-oh-one. Thucydides wrote about it. The Persians did it to the Greeks. Everyone's done it. What makes the modern version different is the complexity of the logistics and the legal architecture around the denial.
Let's step back and define what we're actually talking about here — because "arming a militia" sounds simple, but the reality is anything but.
What we're talking about is proxy arming, or irregular warfare support. And it's distinct from formal military aid to allied nations. When the US sends F-35s to Israel, that's a foreign military sale with a paper trail a mile long, congressional notifications, the whole thing. Proxy arming is designed to be deniable. The whole point is that there's no treaty, no public agreement, often no official acknowledgment at all.
It's the difference between giving your neighbor a signed receipt for a lawnmower and leaving a baseball bat in someone's backyard at 3 a.with a wink.
That's actually not a bad way to put it. And in practice, it's not just handing over rifles. It's a full-spectrum operation. You've got at least four models: direct supply, where the sponsoring government physically delivers weapons. Third-party transfer, where you use an intermediary state or private arms dealer. Financial enablement, where you just provide the money and let the proxy buy their own weapons on the black market. And then training-plus-equipment packages, where you embed personnel who do both.
Each of those has a different risk profile, I'm guessing. The more hands touch the weapons between you and the proxy, the more deniability you have, but the less control.
That's the fundamental tension of this entire practice. Plausible deniability and operational control exist on a seesaw. The more you push down on one, the more the other goes up. If you want total control, you send your own troops. If you want total deniability, you wire money to a cut-out and hope for the best.
Hope is doing a lot of heavy lifting in that sentence.
Hope is not a strategy, but it shows up in a surprising number of these operations. So let me lay out the taxonomy, because the prompt is asking how this works in practice, and the answer depends on which model you're using.
Let's do it.
Model one is direct supply. This is the riskiest in terms of exposure, but it gives you the most control over what gets delivered and to whom. The classic example is the CIA arming the Northern Alliance in Afghanistan in 2001. They flew in weapons, they tracked serial numbers, they did periodic inventory checks. It was basically a military supply chain with extra steps.
At that point, how deniable is it really? If you're running inventory checks, there's a paper trail.
It's barely deniable at all. The deniability there is political, not operational. Everyone knew the US was arming the Northern Alliance. The fiction was that these were indigenous forces fighting their own fight, and the US was just... helping a little. But the weapons had American serial numbers. If a Stinger missile ended up in the wrong hands, it wouldn't take a forensic accountant to figure out where it came from.
Which brings us to model two — the third-party transfer. This is where it gets more interesting.
This is the cut-out system. And it's the bread and butter of intelligence agency tradecraft. The idea is you use an intermediary state, a private arms dealer, or a network of shell companies to obscure the origin of the weapons. The textbook example is Iran-Contra.
Walk me through that one, because it's the gold standard of how these things go sideways.
It's 1985. The Reagan administration wants to fund the Contras in Nicaragua, but Congress passed the Boland Amendment, which explicitly prohibits US military aid to the Contras. So the administration sets up a scheme: sell arms to Iran — which is under an arms embargo — and use the proceeds to fund the Contras. The weapons go through Israel as an intermediary, the money flows through Swiss bank accounts, and the whole thing is run out of the National Security Council basement, basically off the books.
The cut-out was Israel in this case?
Israel was one cut-out. There were also private arms dealers, shell companies, a whole daisy chain. And it worked, in the sense that weapons got to the Contras and the US could technically say it wasn't directly arming them. Until it all collapsed. A cargo plane got shot down over Nicaragua in 1986, a surviving crew member talked, and the whole thing unraveled in about six weeks.
The cut-out works until someone gets caught, and then it fails spectacularly.
That's the thing about cut-outs — they're fragile. Every additional link in the chain is another point of failure. Another person who can be captured, another document that can leak, another bank record that can be subpoenaed. The Iran-Contra affair is the perfect case study because it shows both the ingenuity of the cut-out system and its catastrophic failure mode.
What's model three? Financial enablement, you said.
This is the simplest and arguably the most common. You don't ship weapons at all. You provide money — cash, wire transfers, cryptocurrency now — and let the proxy buy their own weapons on the black market or from corrupt elements of the enemy's military. This gives you maximum deniability because you never touched a weapon. You just funded... Humanitarian aid, maybe. Who's to say?
The "who's to say" defense.
It's remarkably effective, legally speaking. But the downside is you have zero control over what weapons the proxy buys or from whom. You might be funding your proxy, but your proxy might be buying weapons from a dealer who also supplies your enemy. The money winds up in the same ecosystem.
Like giving your teenager cash for "school supplies" and finding out they bought a questionable skateboard and also funded someone else's questionable skateboard.
That's the proxy-arming equivalent, yes. And model four is the training-plus-equipment package, which is the most hands-on. You embed trainers — who are often intelligence officers or special forces — with the proxy force. They provide weapons, but more importantly, they provide knowledge. And knowledge creates dependency.
Dependency is the control mechanism there.
If your proxy needs you to maintain their weapons, to train new recruits, to provide intelligence, to coordinate operations — then they can't easily turn on you without losing all of that. The Phoenix Program in Vietnam is the case study here. From 1968 to 1972, the CIA ran a program that trained and equipped South Vietnamese units to identify and neutralize Viet Cong infrastructure. The trainers were deeply embedded. They controlled the intelligence flow. The proxy units were operationally dependent on the CIA for targeting information.
did not end well, historically speaking.
It's controversial, to put it mildly. But as a model of control-through-dependency, it's instructive. The more the proxy relies on you for things they can't get elsewhere, the less likely they are to turn on you.
Which brings up the other side of this: staged payments and resource dependency as a financial control mechanism.
This is mechanism four in the operational toolkit. Instead of giving a militia a stockpile of weapons upfront, you pay per operation. You provide resources incrementally. You structure it so the proxy needs you to keep the lights on. The "Sons of Iraq" program from 2006 to 2008 is the example here. The US was paying roughly a hundred thousand Sunni tribal fighters a monthly stipend — something like three hundred dollars per fighter — to fight Al-Qaeda in Iraq rather than join the insurgency.
That worked, for a while.
It worked brilliantly, for a while. Violence dropped dramatically. The Awakening Councils, as they were called, were genuinely effective. But the payments were temporary by design. When the Iraqi government took over the program, they stopped paying many of the fighters. By 2010, an estimated twenty percent had rejoined insurgent groups. Some of them later showed up in ISIS.
The control mechanism worked exactly as long as the checks cleared, and not a day longer.
Financial leverage is powerful but brittle. The moment the money stops, the leverage evaporates. And now you've got a hundred thousand trained, armed men with no paycheck and a grudge.
Which is the blowback problem in a nutshell. But before we get to blowback, let's talk about the weapons themselves. How do you supply arms that can't be traced back to you?
This is one of the most operationally interesting parts of the whole practice. There are a few methods. The first is non-standard ammunition — supplying weapons that don't match your own military's inventory. During the Soviet-Afghan war, the CIA deliberately supplied Soviet-designed weapons to the Mujahideen. AK-47s, RPG-7s, DShK heavy machine guns. Weapons that, if captured, would look like they came from Soviet stockpiles or the global black market, not from American armories.
Because if you're the CIA and you hand a Mujahideen fighter an M16, and that M16 gets captured, it's a pretty short investigative path back to Langley.
The CIA bought up Soviet weapons from Egypt, China, and other sources. They also encouraged the Pakistanis to provide weapons from their own Soviet-era stocks. The whole supply chain was designed to produce what's called "plausible provenance." If a weapon is captured, it should tell a story that doesn't include you.
That's a phrase that's going to stick with me.
The second method is removing or altering serial numbers and manufacturer markings. This is more common with small arms. It's not foolproof — forensic ballistics can still match a bullet to a barrel, and metallurgical analysis can identify the source of the steel — but it raises the cost and complexity of attribution.
You're not making it impossible to trace, you're making it expensive and slow.
That's the intelligence game in general. You're not building perfect secrecy, you're building enough friction that by the time anyone can prove what happened, the operation is over and the political moment has passed.
Which loops back to the denial question. The prompt asks whether governments always formally deny these operations. And the answer seems to be yes, but with a very specific logic behind it.
Let's pick that up after we cover one more operational mechanism, because it's directly relevant to the Gaza case. The use of local, clan-based or family-based groups as proxies.
Right, the Abu Shabaab clan specifically.
Clans have several advantages from a sponsor's perspective. They're family-based, which makes them harder to infiltrate — you can't just send a random Hamas operative to join the Abu Shabaab family. They have local legitimacy that an external force can't manufacture. They know the terrain, the social networks, the smuggling routes. And critically, they have their own independent grievances and agendas, which means they don't need to be convinced to oppose Hamas — they already have reasons.
That last part is a double-edged sword, isn't it? The independent agenda that makes them useful also makes them uncontrollable.
That's exactly the paradox. A proxy with no independent agenda is just a mercenary force, and mercenaries go to the highest bidder. A proxy with a strong independent agenda is more reliable in the short term — they're fighting for their own reasons, not just your money — but in the long term, their agenda may diverge from yours. The Abu Shabaab clan has its own local rivalries, its own economic interests, its own political aspirations. Those don't necessarily align with Israeli strategic goals beyond the immediate objective of weakening Hamas.
You're essentially renting someone else's conflict for a while, and hoping the lease doesn't expire at an inconvenient moment.
That's a very Corn way of putting it, but yes. And the history of these arrangements suggests the lease almost always expires at an inconvenient moment.
Alright, let's pivot to the denial question, because that's the other half of what was asked. Do governments always formally deny these operations?
The short answer is yes, almost always. But the more interesting answer is why, and what "denial" actually means in practice.
Because it's not that anyone believes the denial.
Most proxy-arming operations are what intelligence analysts call "open secrets." Everyone in the diplomatic community knows what's happening. The journalists covering the region know. Often the enemy knows. The denial isn't really an attempt to deceive — it's a legal formality.
A legal formality.
Under international law, arming a non-state armed group can constitute a violation of the UN Charter. Article two, paragraph four prohibits the threat or use of force against the territorial integrity or political independence of any state. Arming insurgents inside another country can be construed as an indirect use of force. There's also the principle of non-intervention, which is customary international law. So if you're a government and you're arming a militia inside another country, you need to deny it because acknowledging it would be admitting to a violation of international law.
The denial is less "we didn't do it" and more "we are not formally confessing to a crime.
And this plays out on a spectrum. On one end, you've got the "no comment" — which is not technically a denial, but it's treated as one in diplomatic practice. Then you've got "we have no evidence of that," which is a statement about the state of the evidence, not about the underlying facts. Then there's "those are private citizens acting on their own," which is the most aggressive form of denial — it acknowledges the activity but disclaims state involvement.
All of these are functionally identical from the perspective of the international community.
They serve the same diplomatic purpose, which is to provide legal cover. If you never acknowledge the operation, no one can bring a case against you in the International Court of Justice. No one can invoke the relevant treaty provisions. The denial is the shield.
Let's walk through some historical examples of how these denials have played out, because the pattern is remarkably consistent.
The US and the Contras in Nicaragua is the classic. Throughout the early 1980s, the Reagan administration repeatedly denied providing military aid to the Contras. They used every variant — "no comment," "we have no evidence," "these are private individuals exercising their own judgment." And this continued even as the operation was being run out of the White House basement. The denials only stopped when the Iran-Contra scandal broke and the whole thing became impossible to deny.
The Boland Amendment made it specifically illegal, which meant the denial wasn't just about international law — it was about domestic criminal liability.
That's a crucial distinction. When the arming violates your own country's laws, the denial becomes much more urgent and much more emphatic. The Boland Amendment made it a crime for US officials to provide military aid to the Contras. So the denial wasn't just diplomatic cover — it was a defense against potential prosecution.
What about Israel's historical denials?
Israel has a pattern here too. In the 1990s, Israel was reported to be arming Bosnian Muslims during the Yugoslav wars. The Israeli government denied it consistently. Years later, former officials acknowledged it had happened, but at the time, the denial was absolute. The logic was straightforward: Israel was under a UN arms embargo regarding the conflict, and acknowledging the shipments would have been admitting to a violation.
Russia in Ukraine?
The Donbas is the most recent large-scale example. Starting in 2014, Russia began arming separatist forces in eastern Ukraine. OSCE monitors documented Russian-made 9M133 Kornet anti-tank missiles in separatist hands by July 2014. These are sophisticated weapons that don't just wander off. Russia's denial was comprehensive — they called the separatists "volunteers" and "local self-defense forces" and claimed any Russian weapons in their possession were captured from Ukrainian stockpiles.
Which is the "those are private citizens" variant.
With a twist — "those are private citizens using captured equipment." It's a denial that strains credulity, but again, it's not really about being believed. It's about maintaining the legal fiction that Russia is not a party to the conflict.
That fiction matters because...?
Because if Russia acknowledged it was a party to the conflict, it would trigger a whole set of legal obligations under the Geneva Conventions and open Russia to direct accountability for violations of international humanitarian law. The denial isn't about fooling anyone — it's about managing legal exposure.
The denial is a kind of legal ritual. Everyone knows what's happening, but everyone also understands that the denial has to be made for the system to function.
And the ritual nature of it becomes most visible when the denial collapses. The 2023 Pentagon document leak is a case in point. Leaked documents showed US special forces were training Ukrainian militia units. The official denial lasted about seventy-two hours before the White House acknowledged it. The denial was made because it had to be made — it's the required first move in the diplomatic chess game. But once the evidence was public, the denial served no further purpose and was dropped.
The 2018 Saudi arming of the Yemeni National Resistance?
Riyadh denied it until video evidence emerged of Saudi-supplied weapons in Houthi hands — which is the ultimate failure mode, by the way. Your weapons not only get traced back to you, they get captured by the enemy and displayed on television. At that point, the denial becomes counterproductive because it makes you look both dishonest and incompetent.
There's something darkly comic about that. You deny arming a group, and then the enemy holds a press conference with your weapons on a table.
It happens more often than you'd think. Which brings us to the blowback problem — the question of what happens when the weapons you supplied end up being used against you or your allies.
This is the part of the prompt asking how governments attempt to ensure the arms won't be used against them by groups that are part of the enemy and avowing phony allegiances.
The honest answer is: they can't ensure it. They can only manage the probability. And the historical record is not encouraging.
The Mujahideen is the canonical example.
It's the one everyone points to, and for good reason. The CIA provided an estimated two billion dollars in arms to the Mujahideen in Afghanistan between 1979 and 1989. That included Stinger missiles — shoulder-launched anti-aircraft weapons that could take down Soviet helicopters. They were incredibly effective. But after the Soviets withdrew, the US largely walked away. The weapons stayed in the region. Some of those Stingers ended up in Iran. Others circulated on the black market. And many of the fighters the US had armed and trained eventually became the core of the Taliban and Al-Qaeda.
The US literally spent years and millions of dollars trying to buy back its own Stingers to prevent proliferation.
There was a buyback program in the 1990s. They were offering something like a hundred thousand dollars per missile. They recovered some, but not all. The last known Stinger from that program was accounted for years later. It's the ultimate symbol of blowback — you pay to send the weapons, you pay to get them back, and you still don't get them all.
Like a weapons boomerang that you have to pay for twice.
The Sons of Iraq program I mentioned earlier is another data point. You arm a hundred thousand fighters, the program works, violence drops, everyone declares victory. Then the funding shifts, the fighters are disgruntled, and a significant fraction of them take their weapons and their training to the other side. By the time ISIS swept through Iraq in 2014, some of its most effective commanders were former Sons of Iraq who had been on the US payroll six years earlier.
The control mechanisms — the staged payments, the training dependency, the intelligence control — they all work in the short term, but they don't survive a change in political circumstances.
That's the core insight. Proxy relationships are inherently temporary. They're alliances of convenience. The proxy has its own agenda, and the moment your agenda and theirs diverge, the relationship fractures. The best you can do is structure the relationship so that the fracture, when it comes, does minimal damage.
Which brings us back to Gaza and the Abu Shabaab clan. What's the specific risk profile there?
The risk is acute because Gaza is a closed system. It's a small territory, heavily surveilled, with multiple armed factions operating in close proximity. If Israel arms the Abu Shabaab clan, those weapons are going into an environment where Hamas has its own intelligence apparatus and will be actively trying to infiltrate or co-opt the clan. The question isn't whether some of those weapons will end up in Hamas hands — it's how many and how quickly.
The clan has its own agenda, as you said. What happens when the clan's interests diverge from Israel's?
That's the structural problem. The Abu Shabaab clan is not a pro-Israel organization. It's a clan that opposes Hamas for its own reasons — tribal rivalries, economic disputes, control of smuggling routes. If those reasons change, or if Hamas offers the clan a better deal, the weapons Israel supplied could be turned against Israeli targets literally overnight.
Why take the risk? Why not just use direct military force?
Because direct military force has its own costs — casualties, international condemnation, diplomatic fallout. Proxy arming is attractive precisely because it's cheaper in the short term and comes with a built-in denial mechanism. The calculation is that the short-term benefit of weakening Hamas outweighs the long-term risk of blowback.
Historically, how often does that calculation turn out to be correct?
It depends on your time horizon. In the short term, proxy arming often achieves its tactical objectives. The Mujahideen did help drive the Soviets out of Afghanistan. The Sons of Iraq did reduce violence. The Contras did pressure the Sandinista government. In the long term, the blowback costs almost always exceed the initial benefits, but by then the policymakers who made the decision are usually out of office.
There's a principal-agent problem baked into the structure. The people making the decision won't be around to deal with the consequences.
That's a huge part of why this pattern keeps repeating. The decision to arm a proxy has immediate political benefits and deferred costs. The deferred costs land on someone else's desk.
Let's talk about the counter-infiltration problem specifically. The prompt asks about groups that are part of the enemy and avowing phony allegiances. How do you vet a proxy to make sure they're not already penetrated?
This is one of the hardest problems in intelligence tradecraft. You're trying to build a relationship with a group that operates in enemy territory, often under conditions of extreme secrecy, and you need to verify that they are who they say they are and that they haven't been compromised.
The traditional methods?
There are a few. One is the "bona fides" test — you ask the group to perform an operation that only a genuine opposition group could pull off. Attack a specific target, provide intelligence that can be independently verified, produce a document that only someone with genuine access could have. The problem is that a sophisticated enemy can sacrifice real assets to build a false proxy's credibility.
You ask them to prove themselves, and they do, but the proof might be staged.
Another method is parallel sourcing — you recruit multiple independent proxies and compare their reporting. If one proxy's information consistently contradicts the others, that's a red flag. But this requires having multiple independent networks, which is expensive and time-consuming.
In a place like Gaza, where everyone knows everyone, parallel sourcing might not even be possible. The networks overlap.
That's the clan-based approach's supposed advantage. If you're working with a family structure, the theory is that blood ties provide a natural barrier to infiltration. You can't fake being someone's cousin. But the flip side is that clans have their own internal politics. A faction within the clan might be cooperating with Hamas even if the clan leadership isn't.
You're not just vetting the organization — you're vetting every individual who might have access to the weapons.
That's operationally impossible at any scale. Which is why the control mechanisms we talked about earlier — staged payments, limited stockpiles, training dependency — are really just compensating for the impossibility of perfect vetting.
Alright, so we've covered how the arms get there, the denial ritual, and the blowback problem. Let's pull this together into some practical takeaways, because the prompt is essentially asking for a framework to understand these stories when they appear in the news.
Here's the first insight: the arming of proxies is fundamentally a risk-management problem, not a control problem. You can never fully control a proxy. You can only manage the probability and magnitude of blowback. Every control mechanism — cut-outs, non-standard weapons, staged payments, training dependency — is a way of reducing risk, not eliminating it.
When you read about a government arming a militia, the question isn't "are they in control?" The question is "what's their risk-management strategy, and what are its failure modes?
Second insight: denials are a function of legal exposure, not operational secrecy. The more illegal the operation under domestic or international law, the more emphatic the denial. So when you see a government issuing a very strong denial, the interesting question isn't "are they lying?" — it's "what law would they be breaking if the denial were false?
That's a really useful heuristic. The volume of the denial tells you something about the legal stakes, not the factual ones.
Third insight: the best predictor of whether a proxy will turn on its sponsor is whether the proxy has an independent political agenda. If the proxy is purely mercenary — fighting for money — it'll switch sides when someone offers more money. If the proxy has its own political goals, it'll switch sides when your goals and theirs diverge. Either way, the alliance is temporary.
In the Gaza case, the clans have their own agendas that predate and will outlast any Israeli sponsorship.
That's the structural reality. The Abu Shabaab clan's conflict with Hamas is about local power, not about ideology or alignment with Israel. If the local power calculus changes, the clan's behavior will change. The weapons Israel supplies today could be used in ways Israel never intended tomorrow.
What should a reader actually look for when they see one of these stories? The unnamed officials, the emphatic denials, the leaked reports?
First, look for the legal context. What law would be broken if the denial were false? Is there a UN arms embargo? A domestic law like the Boland Amendment? A treaty obligation? The legal exposure is the story behind the story.
Second, look for the supply chain. Where are the weapons reportedly coming from? If they're Soviet-designed weapons, that's a choice. If they're being routed through a third country, that's a choice. The supply chain tells you how much deniability the sponsor is trying to maintain.
Third, look for the proxy's independent agenda. Who are they? What do they want? What happens when they get it, or when they don't? The proxy's own politics are the best predictor of where the story goes next.
Where does this leave us? The Abu Shabaab case is still active, but it's part of a much older pattern — one that's about to get more complicated.
Because the technology is changing. Drones are getting cheaper. Small arms are proliferating. 3D-printed weapons are becoming more viable. The traditional control mechanisms — dependency on the sponsor for weapons and training — become less effective when a proxy can buy a drone off the shelf and weaponize it with commercially available components.
The seesaw we talked about — deniability versus control — is tilting. Proxies need their sponsors less than they used to, which means sponsors have less leverage.
That's the open question for the next decade of irregular warfare. As the cost of sophisticated weapons drops, does the proxy-arming model become more or less stable? On one hand, cheaper weapons mean less dependency on the sponsor, which means less control. On the other hand, cheaper weapons mean more groups can arm themselves without any sponsor at all, which changes the whole dynamic.
The other frontier is information operations. We've been talking about arming proxies with weapons, but the next evolution is arming their information operations. Disinformation as a force multiplier. A proxy with a Twitter botnet and a weapons stockpile is more dangerous than one with just the stockpile.
That's harder to control than physical weapons. You can track a rifle. Tracking a disinformation campaign is a fundamentally different problem.
The Abu Shabaab story is still unfolding. The test of whether the control mechanisms worked will be whether the weapons supplied to Gaza clans end up in the hands of groups that later target Israel. That's the metric. Everything else is process.
If history is any guide, some of those weapons will end up exactly where they're not supposed to. The question is how many and with what consequences.
Now: Hilbert's daily fun fact.
Hilbert, what do you have for us today?
Hilbert: In the early Renaissance, Honduran mahogany was so acoustically prized that Spanish luthiers would pay smugglers to hollow out logs and float them downriver at night, because the wood's density made it nearly invisible to the harbor inspectors' tapping hammers — a lost script for evading export duties that was rediscovered in a Seville archive in 2003.
That's a new one.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop for the fact and the production. If you want more episodes, you can find us at myweirdprompts.com or on Spotify. We'll be back with another one soon.
Until then, keep an eye on those supply chains.
