Between Oct. 8 and Oct. 11, Ukrainian drones struck three Yandex data centers in the Russian cities of Sasovo, Kaluga, and Vladimir, shutting down two of them completely and damaging a third. At first glance, this seemed like another blow to Russian infrastructure, but the consequences were far more interesting than a simple outage. Yandex is not just a Russian Google. It is the digital skeleton for much of the country’s daily life, providing taxi services, navigation, email, cloud storage, streaming, online payments, and the backend for thousands of businesses. When the facilities went dark, applications stalled, corporate databases became inaccessible, and people across Russia found themselves dealing with frozen screens and endless loading errors. Yandex acknowledged the damage, and Kyiv Post reported that databases, applications, and companies depending on the servers faced serious disruption. Yet the most telling sign emerged in the background: automated traffic from Yandex’s network suddenly plunged. Cloudflare Radar, which tracks internet activity around the world, recorded a sharp fall in machine-driven requests from the company’s infrastructure. These were not human users logging off. They were bots, scripts, and coordinated accounts that had been running inside Yandex’s cloud. The drone strikes had effectively silenced a large part of Russia’s online influence apparatus. It was as if a stadium full of screaming fans had been plugged into a sound system, and someone had just cut the power. The silence was not absolute, but it was noticeable, and it raised a question that has haunted modern politics: how much of what we read online is actually alive? For years, researchers had suspected that a disproportionate share of inflammatory posts in multiple languages were being generated by coordinated networks, but they could not easily prove it, because the accounts seemed to belong to ordinary people. The data-center strikes changed that, at least temporarily, by removing the physical infrastructure needed to run the machines. The result was a kind of accidental experiment: with the servers offline, the chorus of rage suddenly had fewer voices.
President Volodymyr Zelensky drew direct attention to that silence on Monday, Oct. 12, saying there was “less hatred, less discord, and fewer provocations because there were fewer bots,” according to Interfax-Ukraine. It was a striking observation, and one that helped explain why the drone strikes mattered as much as they did. Russian online influence operations are not about a handful of trolls writing angry comments. They are built to manufacture the appearance of public agreement. Moscow’s strategies rely on creating fake identities, cloning news websites, and coordinating thousands of accounts to amplify existing grievances until they look like spontaneous public opinion. The US Justice Department’s 2024 investigation uncovered a sophisticated AI-assisted bot farm developed by a Russian state media employee and backed by an FSB officer, with 968 accounts that often posed as Americans while distributing pro-Russian material. Some of those accounts had convincing profile pictures, biographical details, and posting histories, all generated with the help of artificial intelligence. Another operation, known as Doppelganger, cloned legitimate news sites and invented fictitious profiles to weaken support for Ukraine. These methods are designed to obscure who is actually speaking and to make it seem as though ordinary people are demanding certain policies. When the servers running those operations go dark, the effect on online discourse is immediate. Comment sections become quieter, hashtags lose momentum, and provocative messages disappear. The human tone of the internet returns, at least briefly, because the artificial amplification has been cut off. Zelensky’s words mattered because they framed the drone strikes not only as damage to physical infrastructure, but as an attack on the instruments of manipulation. It is one thing to bomb a factory; it is another to silence a network of fake voices that have spent years poisoning public debate. This is not a simple propaganda model; it is a system that borrows the outward signs of grassroots democracy, and once it stops, people notice the eerie difference between a real conversation and a manufactured one.
The political stakes of these hidden networks are not hypothetical. In Romania, the Constitutional Court annulled the 2024 presidential election after intelligence documents alleged Russian interference and coordinated online promotion of the far-right candidate Calin Georgescu, who opposed military aid to Ukraine and unexpectedly won the first round. The decision was unprecedented in Europe: a national vote was wiped out, not because of fraud at the ballot box, but because of manipulation in the digital sphere. The European Commission later opened proceedings against TikTok, examining risks linked to recommendation systems, coordinated manipulation, and paid political content. In Britain, the issue of Russian meddling has resurfaced as well. Prime Minister Andy Burnham said on Sunday, Oct. 11, that he had received confirmation that Russia interfered in the 2016 Brexit referendum after seeking answers on his first day in Downing Street. His acknowledgment followed a 2020 parliamentary intelligence report that criticized the failure to properly assess possible interference. Burnham did not reveal the underlying evidence or establish exactly how any outside influence affected the result, but the timing of his statement, coming just as Ukrainian drones were silencing bot networks, underscored how fragile public trust has become in the face of digital manipulation. Across Europe, governments are waking up to the fact that the same infrastructure that hosts cloud services can also host armies of fake accounts, and that switching off a data center can have more political impact than a diplomatic protest. The Romanian decision, in particular, showed that elections are no longer won or lost only in ballot boxes; they are also shaped by invisible algorithms, fake engagement, and the coordinated timing of angry comments. When those mechanisms fail, the political ground can shift quickly, and ordinary citizens are left wondering whether their own opinions were ever truly their own.
Some of the most immediate evidence of that change came from Poland and Israel, where observers reported a noticeably quieter internet after the Yandex strikes. In Poland, Russian campaigns had been exploiting disputes over refugees, history, and military aid to intensify anti-Ukrainian sentiment online. An August assessment by the Polish Institute of International Affairs found 364,000 Polish-language posts containing extremely hateful terms directed at Ukrainians in its May-July analysis, generating 1.2 million reactions, comments, and shares. The researchers stressed that Moscow was amplifying existing tensions rather than inventing every hostile post, but the scale was enormous. After the attacks, military commentator Jarosław Wolski described a “de-Russified internet,” saying there were fewer purportedly Polish accounts attacking Ukrainians and fewer purportedly Ukrainian accounts attacking Poles. He also claimed a 56% decline in online support for Germany’s far-right Alternative for Germany party, attaching a graph that showed sharp falls in daily activity. The screenshot offered no identifiable data source or account sample, and activity cannot be equated with voter support, but the observable trend was consistent with the sudden disappearance of coordinated bot traffic. Similar reports came from Israel, where former Ukrainian lawmaker Borislav Bereza said that accounts posting aggressive political comments in Russian and Hebrew had fallen silent. As the country approaches its Oct. 27 election, Ynet reported that the Shin Bet security service had established a dedicated election-security team, with foreign cyber operations and political bots among the threats being monitored. In both countries, the effect was the same: a moment of uncanny calm in comment sections that had been weaponized for years. Some commentators quipped that the internet was beginning to feel human again. The fact that the silence was so noticeable, and so welcome, says a great deal about how much of the online world we live in is actually machine-made.
Now France is bracing for the next test. The country holds pivotal presidential elections next year, and Russia-linked campaigns have already targeted leading figures with fabricated health claims and scandals. Edouard Philippe, Gabriel Attal, and Raphael Glucksmann have all faced false stories, sometimes disguised as reports from established French news outlets. The goal is to destabilize the political landscape and weaken candidates who support Ukraine. As Kyiv Post’s analysis of Marine Le Pen’s presidential bid explained, divisions among pro-Ukraine candidates could benefit her and dramatically reshape French policy toward Kyiv. Le Pen has already ruled out further financial assistance to Ukraine if elected, while leaving room for military assistance and training. That position, if it were to shape French policy, would be a major shift for one of Europe’s most influential countries. The fact that France is preparing for an election at a time when Ukrainian drone strikes are disabling Russian bot infrastructure is both a strategic opportunity and a warning. The attacks on Yandex have given researchers a rare, real-time look at how much of the online conversation is manufactured, but they have also shown that the machinery can be moved, repaired, or rebuilt. No election can assume that a few quiet weeks will last. The French election is likely to become a testing ground for the next generation of influence operations, combining artificial intelligence, fake media outlets, and micro-targeted messaging. The candidates themselves are aware of the danger, but their supporters may not realize how many of the angry voices they argue with online are not their fellow citizens at all. Even a country as digitally alert as France can be caught off guard when the attacks come from multiple directions at once, and with the same credibility as a real news report.
Yandex, meanwhile, is not standing still. The company acknowledged the immediate impact but quickly began arranging for affected customers to move their data and applications to competing cloud providers. It reached agreements with Selectel, K2Cloud, and VK Cloud to accelerate those transfers, offering additional technical teams, simplified onboarding, and flexible financial terms. Yandex also said that rehearsals for the prolonged loss of an entire data-center zone helped engineers switch critical workloads after the attacks. In its own statement, the company said teams were working around the clock to restore services and assess the infrastructure, but it could not give a timetable for bringing the damaged equipment back online. Moving customers elsewhere can restore access before the damaged facilities are repaired, which is good news for Russian businesses and consumers. But the same process carries a darker implication: malign online operations may resume soon, perhaps from new servers, perhaps with a different corporate face. The temporary silence in comment sections across Poland, Israel, and elsewhere was a revealing experiment, showing how much of the internet’s shouting was never human at all. Yet the underlying systems of influence remain in place, waiting for a new home. The drones bought time, and they exposed a hidden layer of modern persuasion, but the struggle over who controls the digital public square is far from over. The quiet after the strikes was a reminder that peace is not merely the absence of bombs, but also the absence of whispers designed to turn neighbors against each other. If the lesson is learned, governments will continue to investigate, expose, and dismantle these networks. If not, the same silence will return one day, not because the bots have disappeared, but because they have learned to hide in better places.