The Punchline Was the Point - Decipon Insights

Introduction

A forged Charlie Hebdo cover doesn’t need a single reader to believe Charlie Hebdo published it. A deepfake of a president kneeling doesn’t need a single viewer to believe he actually knelt. Clown shoes photoshopped onto a wartime leader don’t need to survive a fact-check. They just need to make someone laugh, or sneer, or forward the image to a group chat with no caption at all — because the caption was never the argument. The image was the argument, and its only claim was contempt.

European disinformation researchers have a name for this now: hahaganda. Coined by Latvian academic Solvita Denisa-Liepniece and tracked across a growing body of work by EDMO, EUvsDisinfo, Maldita.es, and StopFake, it describes a specific and until recently under-named current in Russian information operations — propaganda that abandons the pretense of persuasion and goes straight for humiliation. It is worth separating from the disgust-based operations this platform has covered before, because it runs on a different psychological switch, and that switch is harder to flip off once someone’s laughed.

What Hahaganda Actually Claims

Most disinformation, even the crude kind, makes a factual claim you could in principle check: this document is authentic, this soldier said this, this vote was rigged. Hahaganda mostly doesn’t. Per the research documented at EDMO, its content “does not seek to convince the public of its veracity” — it seeks to “undermine the credibility and trust of a specific target… through ridicule and constant humiliation.” A 2024 academic study on humor in Russian foreign policy, cited in the same body of work, describes the register as “aggressive, derogatory,” built to humiliate through exaggeration and implied opposites — take a leader, show him at his least composed, most compromised, most absurd, and let the image do work no argument could.

The primary target has been Volodymyr Zelenskyy, consistently, since Russia’s 2022 invasion of Ukraine. The catalogue of material is specific enough to be checked: manipulated images giving him clown shoes and horns, fabricated Pride-event imagery, AI-generated deepfakes depicting a surrender announcement and a kneeling scene, at least nine forged Charlie Hebdo covers, and fake graffiti physically documented in Madrid, Warsaw, Paris, and Berlin. A parallel track targets Ukrainian soldiers directly — AI videos rendering them as obese and unfit, fabricated “regretful enlistee” narratives, staged prayers-for-survival imagery designed to read as pathetic rather than sympathetic.

None of it argues Zelenskyy is wrong about anything. It argues he’s a joke — and a joke is not a claim you can debunk, because it was never claiming to be true.

The Distribution Model Is the Audience

The content pattern researchers describe starts on Telegram and reaches Facebook, X, and TikTok within hours — not through a bot network pushing it, primarily, but through people. The 2024 study’s framing is precise on this point: hahaganda succeeds when it gets the public to “engage” — to like, comment, and share as active participants, not passive recipients of a message. A forged magazine cover doesn’t need an amplification budget if it’s funny enough that a real person wants their friends to see it. The humiliation isn’t delivered to an audience. It’s completed by one.

That’s a meaningfully different distribution model than a coordinated bot swarm, and it’s why platform moderation built to catch inauthentic amplification can miss it entirely. Nothing about a person genuinely laughing at a mean image and sharing it unprompted looks like a bot network. The account is real, the emotion is real, the share is unpaid and unscripted — the only manufactured element is the image itself, sitting upstream of all of it, engineered by someone who understood that ridicule spreads on its own momentum in a way a policy argument never will.

Why Ridicule Outlasts the Correction

A false factual claim has a natural adversary: the fact. Show the real document, the real timeline, the real quote, and the claim has somewhere to go to die, even if the correction never catches up to the original reach — a well-documented asymmetry, but at least a mechanism exists. Ridicule doesn’t have that adversary, because it was never in the business of claims. You cannot debunk a clown-shoes photo by proving Zelenskyy doesn’t own clown shoes. The image was never asserting that he does. It was asserting that he’s a figure worth mocking, and the only rebuttal to a feeling of contempt is a different feeling, not a fact.

This is what makes it worth naming as a distinct mechanism rather than filing it under the broader banner of “disinformation.” Contempt, once installed, doesn’t require the false premise that produced it to survive scrutiny — the fake Charlie Hebdo cover can be exposed as a forgery in an EUvsDisinfo report, correctly, and the residue of “this guy’s a bit of a joke” can outlive the exposure entirely, because that residue was never load-bearing on the forgery being real. It only needed the forgery to be funny for the length of one scroll.

What This Is Not

This is not the same mechanism this platform documented in The Disgust Operation, which examined Storm-1516 content built around Emmanuel Macron and Friedrich Merz — that operation ran on moral disgust, engineering a visceral revulsion reaction tied to a specific fabricated allegation. Hahaganda runs on contempt and ridicule, a cooler, more social emotion that spreads through shared amusement rather than shared outrage. Both are documented currents inside the same broader Russian information ecosystem; they are not the same tool, and collapsing them into a single “Russian disinfo” bucket would lose the distinction that makes each one detectable in its own way.

This also isn’t a claim of a single unified command structure. The research bodies cataloguing hahaganda — EDMO, EUvsDisinfo, Maldita.es, StopFake, Myth Detector, Delfi, and the EFCSN fact-checker network — are documenting a recurring pattern and a shared vocabulary across incidents, not asserting a single studio produces every clown-shoes image or forged magazine cover. EUvsDisinfo frames these as “active measures” in the traditional Soviet-derived sense — geopolitically motivated, state-linked in orientation — but attribution for any individual meme or graffiti tag documented here rests on pattern and context, the same evidentiary tier this platform has flagged in prior breakdowns of decentralized influence networks. Consistent targeting, consistent register, and cross-border repetition across four capital cities is a strong pattern. It’s not a subpoenaed org chart.

And this isn’t new as a wartime tactic in the abstract — mockery of enemy leaders is as old as propaganda itself. What’s new is the naming, the cross-institutional tracking, and the documentation of why it works on modern distribution: an engagement-driven feed doesn’t distinguish between a share driven by outrage at a lie and a share driven by laughter at a joke. Both register as engagement. Only one required a person to believe anything false.

Key Findings

Hahaganda is propaganda that targets credibility directly, not through a false claim but through ridicule — it does not ask an audience to believe something untrue, per research from EDMO and a 2024 academic study on humor in Russian foreign policy.
The primary sustained target is Volodymyr Zelenskyy, via manipulated images, AI deepfakes (surrender, kneeling), fake graffiti in at least four European capitals, and at least nine forged Charlie Hebdo covers.
A parallel campaign targets ordinary soldiers with AI-generated “obese soldier” content and fabricated regretful-enlistee narratives, aimed at morale rather than leadership credibility.
Distribution runs through genuine audience participation, not primarily bot amplification — content originates on Telegram and spreads via real likes, comments, and shares within hours.
Six-plus fact-checking and disinformation-research organizations across Spain, Ukraine, Georgia, Lithuania, and the EU’s External Action Service are independently tracking the same pattern, cross-referencing it with related operations (Overload, Matryoshka).
The mechanism resists standard debunking because a joke was never a factual claim — exposing a forged image as fake doesn’t reliably remove the contempt the image was built to install.

Implications

The Influence Tactics Protocol treats manipulation as a set of mechanisms, not a genre of content, and hahaganda is a clean case for why that distinction earns its keep. A detection system tuned to flag false claims — fabricated quotes, doctored documents, invented events — will often correctly pass a clown-shoes meme, because the meme isn’t making a claim in the form a fact-checker is built to catch. It’s making a social move: here is someone worth despising. That move doesn’t trip a truth-value check. It trips a completely different reflex, the one that decides who gets taken seriously in a room, and that reflex runs on ridicule as efficiently as it runs on genuine disgrace.

That has a practical consequence for anyone building or relying on detection tools: volume and repetition of a mocking register across a target — four capitals, nine forged covers, a running catalogue of humiliating deepfakes — is itself a signal worth tracking, independent of whether any single piece contains a checkable false statement. The pattern is the evidence, not the individual joke.

For an individual reader, the defense isn’t fact-checking a meme, because there’s often nothing false enough to check. It’s noticing the move underneath the laugh: content whose entire function is making a target smaller, delivered with just enough craft that sharing it feels like commentary rather than complicity. The share is the operation completing itself. Noticing that before forwarding it is the only intervention point that exists once the image has already made you smile.

Conclusion

A joke that isn’t trying to be true can’t be debunked, only recognized. That’s the entire design principle behind hahaganda — it hands a fact-checker nothing to correct, because it never filed a claim in the first place. The forged covers get identified as forgeries, the deepfakes get flagged as synthetic, the graffiti gets photographed and cited in a report, and none of it undoes the specific, cumulative sense of contempt those images were built to leave behind. The mechanism was never in the picture. It was in the laugh.

This article is part of Decipon’s Manipulation Breakdowns series, examining specific influence operations through the Influence Tactics Protocol.

Sources: