The Audience That Wasn't There

The Campaign

For three months before Armenia’s June 7 parliamentary elections, Russia ran an influence operation that researchers described as the most extensive they had seen since Moldova’s 2025 elections. By early May, the Social Design Agency — the Kremlin-funded group behind Storm-1516 and similar operations — had published 343 fake videos targeting Prime Minister Nikol Pashinyan. Armenia had been Storm-1516’s most targeted country in the world between April 2025 and April 2026, according to the Institute for Strategic Dialogue.

On May 11, the UK sanctioned 49 SDA employees — writers, translators, video producers — and named them explicitly as part of a Kremlin-funded operation tasked with installing pro-Moscow figures in Yerevan ahead of the vote. It was a rare public identification of the specific people producing a specific operation in real time, while the operation was still running.

The operation did not work. Pashinyan’s Civil Contract party won with nearly 50 percent of the vote. The pro-Russian Strong Armenia party came in at 23 percent. EU chief Ursula von der Leyen congratulated Pashinyan on the victory. Russia’s foreign ministry spokesperson Maria Zakharova announced that democratic procedures had been “grossly violated.”

The violation she was describing was that Armenian voters had voted the wrong way.

The Specific Lies

The Matryoshka campaign against Armenia was technically sophisticated. It used AI-generated video, networks of fake regional media sites, bot amplification on X, and content laundered through third countries before appearing in Armenian-language information channels. The production quality was high enough to circulate credibly.

The narrative content was more revealing.

The central claim was a fabricated “secret deal” between Pashinyan and French President Emmanuel Macron. The deal, as the fake videos described it, was this: France would provide covert support to Pashinyan’s election campaign; in exchange, after winning, Pashinyan would provoke a military conflict with Russia. The framing served several purposes simultaneously — it introduced France as a foreign election interferer (mirroring legitimate concerns about Russian interference), activated fear of war, and positioned a vote for Pashinyan as a hidden vote for catastrophe.

Other narratives in the campaign included: Pashinyan’s terminal illness and secret medical treatment, NATO instructors covertly stationed in Armenian territory, the existence of “gas chambers on Mount Ararat,” and a general prediction that a Pashinyan victory would mean Armenia would be destroyed in a Russian military response within months.

The Kremlin was not subtle. When the specific claims are listed together, the desperation of the design is visible — fabricated scenarios that individually might be targeted at different fear registers but collectively add up to a wall of catastrophe predictions, each grounded in the premise that Russia is a power that must be deferred to, and that defiance will end in annihilation.

What the Narratives Required

Every influence operation contains an implicit theory of its audience.

The SDA’s Armenia campaign was built on a model of Armenian voters that included several load-bearing assumptions:

That Armenian voters still feared Russia more than they resented it. The “secret deal leads to war with Russia” narrative only functions as deterrence if the audience believes Russia would follow through and that such a war would be catastrophic. This requires Russia to be, in the audience’s emotional register, a formidable and credible power — one whose displeasure is worth structuring your vote to avoid.

That Armenian voters still experienced France and the EU as foreign manipulators. The Macron involvement in the “secret deal” narrative draws on the Storm-1516 playbook that has been working against French and German audiences — the idea that Macron is a conniving globalist willing to use small countries for geopolitical ends. For this to land in Armenia, the audience needs to distrust French motives more than Russian ones.

That Armenia’s relationship with Russia was still ambiguous enough to be contested. The disinformation campaign was designed to create doubt about Pashinyan’s honesty and strategic judgment. That framing assumes voters who are genuinely uncertain, who might be persuaded that a closer look reveals hidden dangers in the pro-European direction.

By June 2026, none of these assumptions were accurate.

The Audience Russia Lost

In September 2023, Azerbaijan completed its reconquest of Nagorno-Karabakh — the territory that had been an Armenian enclave for decades, home to over 100,000 ethnic Armenians. The military operation took approximately 24 hours. Within days, nearly the entire Armenian population of Karabakh fled to Armenia proper in one of the fastest mass displacements in recent memory.

Russia was, formally, Armenia’s security guarantor through the Collective Security Treaty Organization. Russian peacekeepers were stationed in Karabakh. The Kremlin did nothing.

It is difficult to overstate what this meant for the emotional architecture of Armenian domestic politics. The one thing Russia had offered that justified the asymmetric relationship — security — was absent when it was needed at the most catastrophic possible moment. Armenia’s government drew the obvious conclusion and began moving explicitly toward Europe. Pashinyan suspended Armenia’s CSTO membership in early 2024.

This created a specific problem for the SDA’s 2026 operation: the emotional foundation that the campaign’s key narratives required had been destroyed by the same government running the operation.

“Fear a war with Russia” is only a persuasive message to an audience that still regards war with Russia as the worst available outcome. For Armenian voters who had watched Russia stand aside while 100,000 of their people fled a territory Russia had nominally protected, the threat of Russian displeasure had a different valence. The narrative was not threatening. It was a reminder of why the audience had already made its decision.

The “secret deal with Macron” story needed Macron to be frightening and France to be untrustworthy. But the same France that allegedly wanted to use Armenia as a proxy against Russia had been among the loudest European voices supporting Armenian sovereignty and Armenian EU integration. The emotional register the narrative needed didn’t exist.

The gas chambers claim — the most extreme — is worth noting specifically because extreme claims in influence operations usually function as ceiling-setters: if audiences accept even a fraction of the catastrophe framing, the moderate claims seem more plausible by comparison. But when the entire catastrophe framework rests on premises the audience has already rejected, extreme claims don’t function as anchors. They function as signals that the source is not credible.

The Documentation Problem

There is a second reason the operation faced unusual headwinds: it was documented in real time.

Civil society groups, fact-checking organizations, and the Antibot4Navalny collective were actively tracking and debunking the campaign while it ran. The UK sanctions in May named specific individuals. Euronews and Eurasianet published detailed breakdowns of the operation’s structure and specific fabrications weeks before the election.

This created a rare situation: Armenian voters were, to an unusual degree, aware that a large-scale Russian disinformation campaign was targeting their election. The specific “secret deal” videos were fact-checked and their origins traced before they could establish credibility in mainstream media.

Most influence operations depend on obscurity about their provenance. A fabricated video that audiences believe was made by a domestic news outlet lands differently than one audiences know was produced by a Kremlin-funded agency with 49 sanctioned employees. The UK sanctions were not just a diplomatic signal — they functioned as a public prebunking that made the campaign’s outputs harder to accept at face value.

This is not the primary reason the operation failed. Prebunking can slow the spread of content, but it cannot rebuild trust that has been destroyed by events. The fundamental problem was the mismatch between the operation’s audience model and the audience’s actual political reality. But the documentation accelerated the failure and limited the operation’s ability to recover.

After the Vote

Russia’s post-election response completed the playbook.

Maria Zakharova called the election conduct a “gross violation” of democratic norms. Dmitry Peskov, citing what he called “unclear issues” around the vote, said the Kremlin would wait for the final official results before deciding on any Putin–Pashinyan meeting. Translation: we are watching, and recognition is conditional.

This is the final move in the influence operation that failed: replace the pre-election deterrence threat (vote against Russia and face war) with the post-election punishment signal (you voted against Russia, now face consequences). The threat that couldn’t stop the vote is repackaged as a warning about what comes next.

The audience that rejected the pre-election threats is unlikely to find the post-election ones more persuasive. But the function of the post-election messaging is not primarily persuasion. It is face-saving and future-framing — positioning Russia, for its domestic audience and for future operations in other countries, as a power whose displeasure has meaning even when voters chose to ignore it.

Influence Tactics Breakdown

Catastrophe Frame Saturation. Running multiple simultaneous extreme narratives — secret war plots, terminal illness, gas chambers, NATO covert presence — creates a wall of catastrophe predictions that individually might be debunked but collectively establishes an atmosphere of existential threat. The ceiling-setter function works when audiences are uncertain; it fails when the emotional premises the catastrophe requires are already rejected.

Fear-of-Ally-as-Threat Conversion. Converting a target’s new allies (France, EU) into threats by inserting them into fabricated plots — the secret deal where French backing equals war with Russia. This reframes alignment with one’s security partner as secretly dangerous. The tactic requires the audience to distrust the new ally more than the old one; it fails when the old ally has forfeited its credibility through demonstrated inaction.

Laundered Provenance Networks. Using fake regional media sites and third-country amplification to give fabricated content the appearance of local production. The SDA structured content to appear to come from credible Armenian sources before its origins could be traced. The tactic was partially neutralized by real-time tracking and public attribution.

Prebunking as Double-Edged. The UK sanctions in May publicly identified 49 individual SDA employees and named their role in the Armenia operation, creating a named source audience could connect to subsequent suspicious content. Government-level public attribution, when timed ahead of the operation’s key distribution window, functions as a behavioral prebunk — it seeds recognition patterns that lower credibility for the tagged operator’s subsequent content.

Post-Election Punishment Signaling. When a deterrence-based operation fails to prevent an unwanted electoral outcome, converting the threat from deterrence to consequence maintains the credibility of future operations by demonstrating that defiance carries costs, regardless of whether those costs actually materialize. The signal is addressed to future target populations, not the one that just voted.

What This Demonstrates

Influence operations are not universally effective. They are effective under conditions. Those conditions include: an audience emotionally uncertain about the narrative’s premises, a provenance that appears domestic or credible, and no significant competing information environment that has already processed the target narratives.

Armenia in June 2026 had none of those conditions. The audience had processed its uncertainty about Russia through the experience of watching Russia fail them in 2023. The operation’s provenance was publicly documented before the key distribution window. The independent media and civil society environment was actively debunking content in real time.

The SDA ran a technically sophisticated operation that assumed an Armenia that no longer existed. The country that might have been afraid of Russian displeasure and uncertain about EU promises had been replaced, through the Kremlin’s own actions, by a country that had watched its security guarantee dissolve in 24 hours and drawn its own conclusions.

You can run 343 fake videos against an audience that has already made up its mind based on what it has lived through. They still won’t work.

This is worth remembering the next time a successful operation suggests that manipulation is inevitable or that audiences are uniformly susceptible. Operations fail when the emotional terrain they need has been altered by reality. The most reliable form of influence resilience is not media literacy training or prebunking campaigns — though both help. It is the lived experience of watching a power behave in ways that contradict the narratives it needs its audience to believe.

Russia destroyed the precondition for its own influence in Armenia through its conduct in 2023. Then it spent three months making videos about it.

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

Sources: