There is a particular kind of satisfaction when a distant, abstract war suddenly produces something tangible, something you can point to and say: “Look, that was us. That mattered.” Last week, when Ukraine launched drone attacks against three data centres owned by the Russian tech giant Yandex, that feeling swept across the internet. Yandex, after all, is not simply a search engine; it is the backbone of much of Russia’s digital life, often described as “Russia’s Google,” but also so much more. Its cloud services carry everything from email to maps to the kind of behind-the-scenes infrastructure that ordinary people rarely think about. When the company announced that it was struggling to maintain its services, the news felt like a victory. Then came the graphs. Across Telegram, Twitter, and every other platform where people watch the war through the eyes of analysts and armchair generals, users began sharing images that seemed to show a dramatic decline in suspicious online activity by bots. The conclusion was irresistible: the drone strikes had not only dented Yandex’s servers; they had struck a blow against the Russian disinformation machine itself. Even Ukraine’s president, Volodymyr Zelensky, joined in the celebration, saying there was “less hate” online as a result of the strikes. In a war fought as much through screens as through steel, it was a moment of pure catharsis. But catharsis, it turns out, is not the same as truth. As the dust settled, it became clear that this wasn’t the simple success story many people hoped it would be.
The graphs that went viral represented a particular kind of modern hope: that influence operations can be measured, tracked, and neutralised like a missile heading for a radar station. The logic felt straightforward. Yandex data centres are physical places, and if you destroy them, then every service that depends on them begins to flicker. If Russian disinformation bots were running on those services, then the bots would stop. The graphs appeared to show exactly that—a sudden dip in the volume of suspicious accounts and coordinated messaging, almost exactly at the moment the drones struck. For people exhausted by two years of brutal conflict, the visual proof was intoxicating. It was a simple story: attack infrastructure, reduce hate. But numbers, especially those produced in the chaos of a war, do not explain themselves. A graph can show a decline in activity, but it cannot tell you why that decline happened, or what the activity actually was. It cannot tell you whether the bots were truly Russian state trolls or simply ordinary users losing access to their email. It cannot tell you whether the drop was caused by the attacks themselves, by the panic among Yandex engineers, by the platform’s own moderation systems, or by something entirely unrelated. And it cannot tell you whether the bots, if they were bots, have simply moved to another server, another platform, another region of the internet. In the absence of context, a graph becomes a mirror: people see in it what they want to see. And what many people wanted to see was a clean, proportionate, and deeply satisfying response to the endless flood of Russian propaganda.
To untangle what had really happened, the BBC programme More or Less turned to two people who have spent their careers looking behind the surface of online influence. Shayan Sardarizadeh, from BBC Verify, is known for carefully tracing disinformation narratives and separating what can be proven from what is merely plausible. Amil Khan, co-founder and CEO of Valent, has worked on some of the most difficult questions around online manipulation, disinformation, and the ways that hostile actors adapt to pressure. Both were cautious. Both were, in some ways, the antidote to the viral enthusiasm. They pointed out that Yandex is not simply a host for Russian propaganda. It is a massive commercial enterprise with data centres in multiple locations, serving millions of ordinary people—people who woke up one morning to find their email slow, their search engine glitchy, their cloud storage suddenly less reliable. When those services go down, the immediate effect is not a clean reduction in “hate”; it is a messy, human inconvenience. The bots, if they were running on Yandex infrastructure, might indeed have suffered. But the bots are also notoriously adaptable. Russian influence operations have been evolving for years, moving from one platform to another, building robust networks that can survive the loss of a single server or a single provider. A successful influence operation is not a static machine sitting in a data centre; it is a living, breathing ecosystem, with redundancies, backup systems, and, increasingly, automated tools that can be spun up in a matter of hours. Even if the attacks did disrupt some bot activity, the effect may have been barely more than a flicker in the larger campaign of deception.
This is where the human side of the story becomes more complicated. The desire to believe in the “less hate” narrative is understandable, especially for a country under invasion. Ukraine has been fighting not only military forces but also a relentless information war, one that has at times been amplified by the very platforms where those graphs circulated. When Zelensky says there is less hate online, he is giving voice to a people desperate for evidence that the invisible enemy can be hurt, that the propaganda machine is not invincible. That is a deeply human need. But it is also a dangerous one, because when leaders and citizens cling to unverified statistics, they risk building strategy on hope rather than evidence. Amil Khan’s expertise suggests that influence operations are rarely defeated by a single blow, no matter how spectacular. They mutate. They migrate. They find new homes on platforms that are less moderated, new technologies that are less understood, new narratives that are designed to exploit the very confusion that follows an attack. The Russian disinformation ecosystem is not located in three data centres; it is located in the minds of millions of people, and it is sustained by a constant process of adaptation and reinvention. The drone attacks were not meaningless—they caused real damage to an important company and sent a message about Ukraine’s ability to strike deep behind enemy lines—but they did not, by themselves, dismantle the propaganda machine. To claim otherwise is to misunderstand the nature of the beast.
The story of the Yandex graphs is, in the end, a story about how we process numbers in the age of social media. We are all, in some sense, amateur data analysts now. We scroll through screenshots and charts, and we make snap judgments about what they mean. A graph showing a drop in bot activity feels like evidence because it has a shape, a line, a clear before-and-after story. But a graph is not a fact; it is an argument. It is a way of presenting some data while hiding others. It does not tell you where the data came from, how it was collected, or what assumptions were baked into its creation. It does not tell you whether the people who made the graph were acting in good faith, or whether they were simply so eager for good news that they forgot to ask the hard questions. The More or Less team, led by presenter Lizzy McNeill, asked those hard questions. They asked what the graphs were actually measuring, what the limits of the data were, and whether the conclusions being drawn could survive contact with reality. Their answer was not a simple yes or no, but a warning: be suspicious of the story that feels too clean, especially in times of war. The truth is usually messier, more layered, and more uncomfortable. That does not mean the drone attacks were a failure. It means that the full picture remains incomplete, and that filling in the gaps requires patience, humility, and a willingness to admit that we do not yet know what we thought we knew.
So where does that leave us? It leaves us with a military action that may have had real effects, but not necessarily the effects that were claimed. It leaves us with a Russian tech giant struggling to maintain services, and with millions of ordinary users caught in the middle. It leaves us with a community of observers who, in their hunger for a decisive blow against disinformation, took a plausible story and turned it into a certainty. It leaves us, too, with an important lesson about the nature of propaganda. Propaganda is not just the lies that are told; it is also the comfortable truths we tell ourselves. The idea that a single drone strike can silence the machine that generates so much online hatred is, in its own way, a kind of wishful thinking. It converts a sprawling, adaptive, and deeply human enemy into a mechanical target that can be destroyed with precision weapons. It makes the world simpler than it is. The experts who spoke to More or Less did not offer a neat alternative narrative; they offered something more valuable. They offered a method: check the numbers, question the graphs, consider the alternatives, and do not confuse a temporary service outage with a lasting strategic victory. And they reminded us that the fight against disinformation is not a battle that can be won once and for all. It is a continuous, exhausting, and essential process—one that requires all of us to be more sceptical of the stories that make us feel good, not less. If you have seen a number or a graph that you think deserves that kind of scrutiny, the More or Less team would like to hear from you. Because the next time a shape appears on your screen promising a clean victory, you may want to pause and ask: what else could this be? The answer might surprise you. It might even be more interesting than the graph itself. Credits: Presenter: Lizzy McNeill, Producer: Nathan Gower, Series Producer: Tom Colls, Production Co-ordinator: Maria Ogundele, Sound Mix: James Beard, Editor: Richard Vadon.