The problem with a category like this is the name itself does half the conspiracy work for you. Covert operations to degrade an enemy's society from within. You say that out loud and you've already sold the book.
And the book probably has a redacted cover and a foreword nobody can verify.
Daniel's prompt this week is asking us to separate the thing that's real from the thing that's a publishing imprint. He wants the documented cases where governments have covertly used misinformation and other tactics to degrade an enemy's internal cohesion as a weapon of war. Stoking ethnic division, political division, religious division, class division, whatever fault line is already there, and turning it into the actual battlefield. And he wants us to be honest about which dramatic claims in this category are unfounded or exaggerated, versus which ones have at least a kernel of truth that checks out in archives.
Which is the harder question, honestly. Because the real operations and the myths feed each other. Once you've got one documented case, every unsourced story about the same tactic suddenly borrows credibility from it.
So let's walk through the mechanism first, then the cases that hold up under scrutiny, and then the part where the whole thing tips into unfalsifiable territory.
And the mechanism matters because this isn't ordinary propaganda. Propaganda is broadcasting your own message. This is different. The objective isn't to convince the target population of something. It's to get the target population to tear itself apart.
Right. You're not trying to defeat the enemy's army. You're trying to make the enemy's society ungovernable, or at least distracted enough that it can't act coherently.
And the key structural feature, the thing that keeps showing up in every documented case, is that the operation works best when it amplifies divisions that already exist. You don't invent a grievance. You find one that's already there, maybe half-settled, maybe simmering, and you pour fuel on it.
If you invent a grievance from nothing, the target society's immune system kicks in. People say, wait, that doesn't sound like us. But if you take a real argument that's already happening and you sharpen it, fund one side of it, publish forgeries that confirm what one side already suspects, then it looks organic. Because it mostly is organic. You're just turning up the heat.
And that's also why the historical record is so uneven. These operations are designed to be deniable. The whole point is that the target can't prove what happened. So some cases are documented because archives opened after the Cold War. Some are known through defectors. Some are known because a parliamentary commission got access to files. And some things that definitely happened will never be provable because the paper trail was destroyed.
Which is the exact gap where conspiracy theories move in.
When the evidence is classified, the absence of evidence becomes its own kind of evidence for people who want to believe the bigger version.
So let's start with the clearest institutionalized framework for this kind of warfare, because it wasn't a rogue operation. It was a standing bureaucracy.
The Soviet concept of active measures. Aktivnyye meropriyatiya. This was a formal category of operations run by the KGB and its allied services, and it included disinformation, forgeries, influence operations, political manipulation, all of it coordinated as policy. Not a one-off. A department.
And the important thing about active measures is that they were considered a legitimate peacetime tool. This wasn't wartime emergency stuff. This was the KGB's day job, running influence operations against Western societies continuously.
The canonical case, and the one that best illustrates the mechanism, is Operation INFEKTION. This was a Soviet campaign to spread the claim that the United States invented HIV/AIDS as a biological weapon. Seeded in nineteen eighty-three through a Soviet-funded newspaper in India, then amplified through East German and Soviet outlets.
The Patriot. That was the paper, wasn't it?
The Patriot, yes. A small-circulation English-language paper in New Delhi. And the initial article cited an anonymous American scientist and claimed the virus was engineered at Fort Detrick. From there it got picked up by Soviet state media, by East German outlets, and then it started circulating internationally.
And here's the part that makes it a masterclass in the mechanism. The claim didn't need to convince everyone. It needed to create enough ambiguity that people who already had reasons to distrust the United States now had a pseudo-factual anchor for that distrust.
You didn't have to believe the whole thing. You just had to think, well, I'm not sure, but it's plausible enough, and there's a newspaper article, and there's a scientist mentioned. And once that seed is in, it's very hard to remove. The claim persisted for years. It showed up in mainstream international media. It got cited in some scientific circles. There were African leaders repeating it. And it's still circulating in some places today.
And the Soviets didn't have to do the convincing themselves. The target's own media ecosystem did the work. Once the story was out there, legitimate journalists, researchers, and activists picked it up and repeated it without knowing the origin.
That's the delivery mechanism. The operation weaponizes the target's own information ecosystem. The Soviet Union didn't need to stand up a radio station and say, the Americans invented AIDS. They needed one article in an obscure Indian paper, and then the global information system did the amplification for free.
And the target's existing anti-American narratives did the rest. The story fit a pre-existing worldview, so it got traction in places where that worldview was already strong.
Now, the second case I want to talk about is different in kind but identical in logic. The strategy of tension in Italy during the Cold War.
This one's messier, because the documented parts and the contested parts sit right next to each other.
Right. So the background is that Italy in the nineteen sixties and seventies had a very strong Communist Party, and there was genuine fear, both in Italy and among Western allies, that the left could come to power through elections. And during this period, Italy experienced a wave of bombings and political violence. The strategy of tension was the idea that these attacks were being used to push public opinion toward authoritarianism and away from the left.
And the documented part is that there were covert stay-behind networks, the most famous being Gladio, that existed across Western Europe. These were paramilitary networks set up during the early Cold War, originally to resist a Soviet invasion, but in Italy they appear to have been involved in domestic politics.
Gladio was exposed in nineteen ninety, and the Italian parliament launched an investigation. The existence of the network is confirmed. That's not in dispute. The question is how directly specific bombings were directed by intelligence services, and that part remains contested in Italian courts and historiography.
So you've got confirmed covert networks, confirmed political violence, and a confirmed pattern of the violence being used to justify crackdowns. But the direct line from, say, a specific intelligence officer to a specific bomb is much harder to draw.
And that's the honest version. Some accounts make it sound like every bombing in Italy for twenty years was a CIA operation. That's not what the evidence shows. But the broader pattern, the use of violence and fear to shift the political center of gravity, that's documented enough that an Italian parliamentary commission took it seriously.
The through-line between INFEKTION and the strategy of tension is the same. Use the target's own media, institutions, and existing grievances as the delivery mechanism. Don't import a foreign narrative wholesale. Make the operation look like it's coming from inside the society.
In the INFEKTION case, the delivery mechanism was the global press and pre-existing anti-American sentiment. In the Italian case, the delivery mechanism was domestic far-right groups and the existing fear of communist electoral success. The operation doesn't create the division. It rides it.
And the deniability comes from the fact that the people doing the visible work don't know who's ultimately directing them. The Indian newspaper editor who ran the AIDS story probably thought he had a scoop. The Italian neo-fascists who set bombs probably thought they were acting on their own ideology. And they may have been.
That's the part that makes these operations so hard to prove after the fact. You can't find the memo that says, go bomb this piazza. Because the memo probably didn't exist. The operation worked by creating conditions where someone else would do it.
So that's the institutionalized version. A standing apparatus for active measures. But the logic didn't disappear when the Soviet Union did. It just got cheaper and faster.
And the modern version is structurally different in ways that matter. The target society's own social media platforms become the amplification layer. So the operation is cheaper, faster, and much harder to attribute than anything the KGB was doing in nineteen eighty-three.
The clearest contemporary documented case is Chinese influence operations targeting Taiwan's elections. We're not talking about speculation here. This is documented, coordinated inauthentic behavior, disinformation campaigns, pressure on Taiwan's media and political discourse.
And the objective is the same as the Cold War cases. Not to defeat Taiwan militarily. To degrade its internal cohesion. To make its politics more fractured, more suspicious, more paralyzed.
The mechanism is almost identical to INFEKTION, just with different infrastructure. You don't need a Soviet-funded newspaper in New Delhi anymore. You need a few hundred fake accounts, some coordinated hashtags, some manipulated screenshots, and the platform's own recommendation algorithms do the amplification.
And the cost is effectively zero compared to the Cold War version. The KGB had to maintain front organizations, pay journalists, print newspapers, ship copies around the world. Now you can run a comparable operation from a single office with a budget in the tens of thousands of dollars.
The attribution problem is also worse. In the Cold War, you could eventually trace the money or the paper trail. Now the operation is distributed across platforms, across jurisdictions, and the people running it are using infrastructure that's designed to be anonymous.
So those are real. The documented cases are real. But here's where it gets tricky. Once you know this kind of operation exists, it becomes very tempting to see it everywhere.
That's the second half of Daniel's prompt. Which claims are unfounded or exaggerated? Because the category is a magnet.
The pattern is predictable. A genuine case gets documented. Say, INFEKTION. Then every subsequent claim about disinformation borrows the template. Any time there's a divisive narrative circulating, someone says, this is INFEKTION again. This is foreign-engineered.
Most of the time, it isn't. Most of the time, the divisive narrative is homegrown. It's the target society's own politics, its own media, its own people doing what they do.
The specific failure mode here is that attributing domestic polarization to foreign actors can become a form of denial. It lets a society avoid examining its own internal drivers of division. If you can blame the Chinese or the Russians or whoever for your political dysfunction, you don't have to do the harder work of asking why your own institutions are so fragile.
That's the comfortable version. The uncomfortable version is that a lot of the division is real, it's local, and it would exist with or without foreign interference.
The asymmetry problem compounds this. The same tactic, stoking division, is used by many state and non-state actors. So attribution is difficult. And premature attribution can be as damaging as the operation itself.
If you publicly blame a foreign actor and you're wrong, you've just handed that actor a propaganda victory and you've discredited your own institutions. If you're right but you can't prove it, you've still created a controversy that the foreign actor can exploit.
The knock-on effect is the one I keep thinking about. The existence of these operations creates a permanent epistemic tax. Citizens can never be fully certain whether a given divisive narrative is organic or engineered. That uncertainty itself degrades social trust.
Which means the operation succeeds even when it fails. Even if nobody believes the specific disinformation, the fact that the disinformation existed, the fact that people now have to ask, is this real or is this a Russian operation, that question itself is corrosive.
It's the perfect weapon for an adversary who wants to degrade your society. It doesn't need to win arguments. It just needs to make every argument suspect.
The two problems feed each other. The real operations make the exaggerated claims seem plausible. The exaggerated claims make it harder to take the real operations seriously, because people see the category as tainted.
You end up in a situation where the intelligence agencies are warning about real threats, but the public has heard so many overblown claims that they tune out. Or the opposite, where the public is so primed to see foreign interference that they attribute every domestic dispute to it.
Both are failures. And both are exploitable.
The thing I want to stress is that the low-level participants in these operations often have no idea what they're part of. That's not a defense of them. It's a structural fact. The operation is designed so that the people doing the visible work are cut off from the people directing it.
Which is what makes it deniable. If the newspaper editor doesn't know he's running a Soviet operation, he can't confess to it. If the social media user doesn't know they're amplifying a coordinated campaign, they can't be flipped.
That's the part most coverage gets wrong. It imagines these operations as a bunch of spies in a room pulling strings on a puppet. The reality is much more diffuse. It's a set of conditions, a set of incentives, a set of narratives that are seeded and then left to grow on their own.
The KGB didn't convince millions of people that America invented AIDS. They convinced a few editors, a few journalists, a few activists, and then the existing information ecosystem did the rest.
The Chinese influence operations in Taiwan work the same way. They don't need to convince Taiwanese voters of anything. They need to amplify the most divisive voices, fund the most extreme positions, and make the political conversation so toxic that nobody can govern.
The goal isn't a specific outcome. It's dysfunction.
Which brings me to the part of Daniel's question that's hardest to answer. How do you tell the real operations from the myths?
The honest answer is that you often can't, not in real time. The evidence comes later, if it comes at all. The INFEKTION campaign wasn't fully documented until the Soviet archives opened. The Gladio network wasn't exposed until nineteen ninety, decades after it was set up.
The calibration problem is real. You can't ignore the threat, because the documented cases show it's real and it's persistent. But you can't see it everywhere, because most of what you're seeing is probably domestic.
The cost of getting it wrong in either direction is high. Ignore it and you're vulnerable. See it everywhere and you've paralyzed yourself.
I don't have a clean answer to that. I think the best you can do is maintain a kind of epistemic discipline. Demand evidence. Be specific about what's documented and what's alleged. And resist the temptation to use foreign interference as an explanation for domestic failures.
Which is easier said than done, especially when the foreign interference is real.
Especially then.
Hilbert: You're both right about the low-level people.
Say more.
Hilbert: I worked a translation job in the late eighties. Small publisher in West Germany, did current affairs paperbacks. One of the books they gave me was a translated Soviet-sourced tract about AIDS. The author was a name nobody could verify. The manuscript appeared fast, like it was waiting for us. And the editorial instructions were strange. Very specific about certain phrases. Very insistent that we keep the phrase biological warfare program in the subtitle.
You think you were part of INFEKTION.
Hilbert: I didn't think anything at the time. It was a translation gig. Paid by the page. But the book was exactly the claim. The Americans created it in a lab. Fort Detrick. The whole thing. And the publisher was a mail-drop. I found that out years later.
A mail-drop.
Hilbert: The address on the copyright page was a post office box in a town that didn't have a publishing house. I had a copy of the book for years. Never opened it after the job was done. Then a few years ago I read about INFEKTION and I thought, that's the manuscript.
Did you keep the copy?
Hilbert: It's in a box somewhere. I looked at the copyright page once, after I learned about the campaign. The publisher's address was a mail-drop. The translator credit was a name I'd never heard. And the introduction was signed by a doctor who didn't exist. I checked.
You were a low-level participant in a Soviet active measure.
Hilbert: I was a guy who translated a book. That's the whole point. I didn't know anything. The publisher didn't know anything. The bookshop that sold it didn't know anything. The person who bought it and read it didn't know anything. But the book existed, and it said the thing, and it sat on shelves in West Germany with a fake author and a fake publisher and a real ISBN.
That's the operation. It's not a spy in a room. It's a book on a shelf.
Hilbert: It's a book on a shelf that someone paid me four hundred marks to translate in a week. I thought I was doing a rush job for a cheap publisher. I was doing a rush job for a cheap publisher. That's what makes it work.
The deniability is built into the labor structure. The translator doesn't know. The editor doesn't know. The typesetter doesn't know. Each person does one small thing.
Hilbert: If you ask any of us, we'd say we did a job. Which is true. That's the whole thing. The operation is just a lot of people doing jobs.
The people who read the book, they think they're reading a book. Which they are.
Hilbert: A book with a fake author and a fake publisher and a real argument. And the argument did what it was supposed to do. It gave people who already distrusted the Americans a reason to keep distrusting them. It didn't need to convince anyone new.
It needed to be citable.
Hilbert: It was citable. That was the point of the editorial instructions. Make it citable. Make the footnotes look real. Make the author sound like a scientist. So that when someone wanted to say, the Americans invented AIDS, they could point to a book.
Now we're back to the epistemic tax. The book existed. The claim was false. But the citation persists.
Hilbert: The citation persists. I saw the claim in a documentary a few years ago. They cited a book. I didn't check if it was the one I translated. I didn't want to know.
I think that's the part people don't get. The operation doesn't end. The book stays on the shelf. The citation stays in the literature. The claim stays in circulation. And the people who did the work go on with their lives.
The people who read the claim have no way of knowing that the translator was a guy in West Germany who thought he was doing a rush job.
Hilbert: I was doing a rush job. That's all I was doing. And thirty years later I'm sitting here realizing I was part of a Soviet active measure. The thing you're describing, I did it. But I did it.
That's the mechanism in one anecdote.
Hilbert: It's the mechanism in one book. And the book is still in a box somewhere.
The question Daniel asked, the calibration question, it doesn't get easier with time. Because the evidence arrives decades late, if it arrives at all. And by the time it arrives, the damage is done.
The damage is done and the citation persists. That's the thing I keep coming back to. The operation succeeds on a delay. The book comes out in nineteen eighty-seven. The claim circulates for decades. The citation shows up in a documentary in twenty twenty. And the translator only realizes what he did thirty years later.
What's the open question? If the documented cases are real but the category is also a magnet for unfounded claims, how should a society calibrate its response?
I don't think there's a formula. I think the best you can do is insist on evidence, be specific about what's documented, and resist the temptation to use foreign interference as an excuse for domestic failures.
As generative AI lowers the cost of producing plausible disinformation, the attribution problem gets harder. The epistemic tax gets higher. The book on the shelf gets easier to produce and harder to trace.
The unresolved tension is the point. The real operations are serious. The exaggerated claims are also serious, because they poison the well. And both problems are getting worse.
This has been My Weird Prompts. Thanks to our producer, Hilbert Flumingtop, for keeping the show running.
If you enjoyed this episode, leave us a review wherever you listen. It helps more than you'd think.
We'll be back soon.
