Introduction
On Aug. 6, the Ukrainian army started an incursion into Russia's Kursk region. As of Aug. 19, Ukraine controls more than 1000 km2, disrupting Russian logistics and supply lines. The rapid start of the operation caught the Russian army unprepared. Almost two weeks later, the Ukrainian military still successfully maintains pressure in the region.
JOIN US ON TELEGRAM
Follow our coverage of the war on the @Kyivpost_official.
The event was unexpected for the Russian army, media and citizens, creating a space for uncoordinated narratives. For example, local administrations and pro-Kremlin media mostly reported about the evacuation, humanitarian aid to the locals and “successful” defence against the attackers. On the other hand, general population pointed out the lack of support from the state and were irritated because of the inability of the Russian army to prevent and counter the Ukrainian offensive.
We have explored key narratives in Russian social media and news outlets, extracting 1,216,094 documents related to three topics: the incursion, mobilisation in Russia, and the Russians’ attitude towards the war with Ukraine. Here is what we found out.
Executive Summary
The incursion immediately became a prime topic in the Russian internet. We identified a 196 percent increase of the amount of content about the war compared to the week before the incursion started. In total, we have collected more than 470,000 documents about the incursion from 800 various sources, predominantly social media like VK, Telegram, Odnoklassniki and Youtube. The peak in the number of the publications was on Aug. 8 with over 128,000 publications during one day.
Estonia Moves to Ban Russian and Belarusian Residents from Voting in Local Elections
The Russian propaganda narratives related to the incursion have been unusually uncoordinated. Different war correspondents raised panic about the losses of the Russian army while the pro-Kremlin media outlets downplayed the scale of the event, calling it a “provocation.” This anomaly shows that the incursion was unexpected for both Russian authorities and propagandists.
The events in Kursk seem to have affected Russian perspectives of the war. The sentiment towards the war has hit the lowest mark in the last 70 days, falling down to -0.47. This decrease has two components: firstly, there have been fewer cheerful publications about the war. Secondly, there were more grievances compared to the previous two months. The grievances featured blaming the Russian authorities and general panic regarding the incursion.
We have also identified a growing concern about the potential new wave of mobilisation in Russia in response to the incursion. During the first week of the incursion, approximately 39 percent of the publications about mobilisation mentioned the Kursk incursion – authors and commenters discussed if the new wave of mobilisation was to come. The concerns among the Russian population made a Russian MP Viktor Zavarzin refute the information that the army plans to mobilise additional soldiers due to the incursion. Also, among the most popular comments in VK was a discussion of Russian prisoners of war and their fate – this topic accounted for 11 percent of the comments related to the incursion.
Russian leadership seems to reflect citizens’ concern about the mobilisation. The Kremlin has not instrumentalised the incursion as a means to mobilize the general population against the “existential threat” to the state. We suggest that the Russian government understands the sociopolitical risks of a new wave of mobilisation and fears the potential consequences related to it.
Local officials employed the tactics of relatively neutral communication: informing about the evacuation (25 percent of the content), behaviour during the shellings (24 percent), and bomb shelters (18 percent). At the same time, the general population didn’t agree with the official line, with 12 percent of the comments from Russian citizens directly blaming Russian authorities (the quotes are mentioned in the “General population” chapter). One of the key topics was the critique of the poor preparation of the Russian army. Also, people expressed concerns about the lack of social support for the affected population, like the compensations for the destroyed property for the Kursk region population. Most of the content about the incursion was dedicated to the coverage and analysis of the combat operations, with a few prominent ideological narratives.
Pro-Kremlin media outlets claimed that the incursion is pointless because Ukraine is losing its territories in Donbas. This argument featured in 8 percent of their publications related to the incursion. This number is significant since more than a half of the content from pro-Kremlin outlets consists of relatively neutral news reporting. They stated that the operation did not make sense for Ukraine, using the Russian army’s advancement in parts of the Donetsk region as the key argument. The main idea behind this narrative is “we will win no matter what, this small incursion will not get in our way”. Propaganda outlets devoted over 20 percent of their content to covering the affected civilian population. Their coverage ranged from highlighting humanitarian and social support for affected populations to using local stories as a tool of demonising the Ukrainian army.
Pro-Russian war correspondents pushed the narrative about NATO being the main actor in preparing and even executing the operation. This narrative was among the most popular ones, accounting for 14 percent of war correspondents’ content on this topic. They claimed that the Western powers are clearly behind the incursion, from supplying the weapons to sharing intelligence data and planning the whole operation.
All actors shared a concern about the potential combat operations near Kursk Nuclear Power Plant (hereafter – Kursk NPP). This topic was among the top-5 most popular ones across different sources, accounting for 12 percent of the publications related to the incursion in pro-Kremlin media and 6 percent in the comments of general population.
Methodology
For this research, we sourced data from around 752 channels, including the most popular social media in Russia such as vk.com, telegram, youtube.com, and odnoklassniki.ru, and media outlets such as news.ru, ria.ru, rbc.ru. The data collection period spanned from Aug. 5 to Aug. 12, resulting in the aggregation of 472,419 documents. Additionally, we have collected more than 700,000 documents on broader topics about war and mobilisation. A "document" in this context refers to posts, comments, and articles that we extracted and analysed from Russian social networks and other sources.
Data was collected using the 240 keyword pairs, such as “Суджа ВСУ” (“Sudzha AFU”), “Курск солдаты” (“Kursk soldiers”), “Рыльск эвакуация” (“Rylsk evacuation”).
Additionally, we collected data about two other topics in the same time frame – attitude towards the war with Ukraine (646,245 documents from 1,137 channels) and attitude towards mobilisation (97,430 documents from 612 channels). Combined, we collected 734,675 documents on the two topics.
We also separately collected posts from 210 Telegram channels, including state and pro-Kremlin media, war correspondents, and local Kursk channels, along with comments on them, totalling 20,000 messages. Using the VK API, we gathered 35, 000 posts based on a set of keywords and scraped 53,000 comments on these posts.
We have also split the types of the sources we use, dealing with the following types: pro-Kremlin media outlets; war correspondents; local officials; local media groups; and the general population.
Sentiment towards the incursion was identified using a machine learning model. To measure the attitude towards the topic we calculated the proportion of positive comments to negative ones. As a result, the sentiment expressed in the posts, comments, and articles is normalised on the scale from -1 to 1. A value of -1 indicates the most negative sentiment, while a value of 1 indicates the most positive sentiment.
We also employed topic modelling, a machine-learning technique for discovering common patterns in large collections of texts. This method facilitated the identification of primary document clusters (document groups based on the similarity of their contents). We also calculated sentiment associated with each cluster as a mean sentiment value of the associated documents. Short summaries were generated for each cluster with a GPT 3.5-Turbo language model. A combination of these techniques provided our research team with a high-level overview of online discourse around all four topics.
Main narratives about the incursion
We separately analysed sources of different types to understand how the narratives of Russian pro-government media, war correspondents, local authorities, and general population differ.
Pro-Kremlin media
During the first week of the incursion, pro-Kremlin media mostly reported on the combat operations (28 percent), analysed the strategy of the Ukrainian army (27 percent), and stressed out the casualties of the Ukrainian army (9 percent). However, among the most popular topics, making up 22 percent of the sample, was highlighting support for the affected civilian population, covering the displacement of local residents and providing humanitarian aid to them.
Moreover, 4 percent of pro-Kremlin media content was dedicated to the casualties among the civilian population – this sub-topic is highlighted within 22 percent about civilians. Such materials unanimously blame the Ukrainian army as the reason behind the casualties.
“The man was hospitalised with a shrapnel wound to the chest. The woman, who also received a shrapnel wound to the leg, was taken to the hospital by ambulance.”
“Acting regional governor Alexey Smirnov reported that shelling of the city of Sudzha injured five people, including three children”
“"If you raise your head again, we'll shoot to kill": A resident of Kursk Oblast about a hike through militant-occupied territory”
“One person suffered from barotrauma. Medical assistance was provided at the scene, but he refused hospitalisation.”
The pro-Kremlin outlets dedicated 12 percent of the content to the issue of Kursk NPP and potential consequences of active combat near it or even its occupation by Ukraine. The smaller topics include reporting on the wounding of a Russian war correspondent Evgeniy Poddubnyy (3 percent). Also, around 5 percent of the content was dedicated to demonising Ukraine – in this context, the media exploited the “first hand” stories told by the local residents who evacuated. However, such a demonisation might be featured in other topics, like the one about the Kursk NPP or analysis of combat operations. In this particular topic we only captured a particular cluster of “first hand” civilian stories used to describe the Ukrainian army as war criminals. This content fueled xenophobia towards Ukraine in Odnoklassniki – 14 percent of the content in the social media regarding the incursion consisted of insulting Ukrainians with derogatory language.
Two percent of the publications pointed out that the US and other Western countries are directly engaged in the incursion: “USA stand behind the terrorist attack on the Kursk region,” “Western media are clumsily trying to convince their audience that Kyiv's actions came as a surprise to the United States and other allies. But based on publicly available information, it can be concluded that the terrorist attack in the Kursk region was planned and carried out by the United States”
Also, 1 percent of the content featured the fate of the gas measuring station in Sudzha: “Control over the gas metering was important for Ukraine, as it could create favourable conditions for negotiations. The pumping of gas through Sudzha was indeed reduced and stopped”.
We decided to explore the outreach of pro-Kremlin media and how its content resonated with its audience. To do this, we calculated the views per post and forwards/shares per post for each topic. The leading topics are the same both in terms of views and forwards: Evgeniy Poddunbyy, combat operations coverage, and the content about civilians. On the other hand, the audience practically did not engage with the content that demonises Ukraine. The topic about Kursk NPP has not received much engagement, either.
War correspondents
We separated war correspondents from pro-Kremlin outlets, because they might express different views despite being affiliated with the government and supporting the Russian invasion. Moreover, they have relatively big audiences, with one of the most popular war correspondents, the mentioned Evgeniy Poddubnyy, having 781,000 subscribers on his Telegram channel. This makes them influential actors in the Russian information space. In total, we gathered 1,095 publications from 39 war correspondents.
Russian war correspondents are mainly focused on reporting on-site – 41 percent of the discussions are related to coverage of combat operations in the region. They often refer to NATO as a direct participant in the Kursk Incursion (almost 14 percent of the posts), downplaying the role of Ukrainian military command and focusing on its strong dependency not only on Western weapons, but on their tactics and strategies: “NATO's hand is 100 percent visible in the offensive in terms of support with intelligence data and in the planning of the operation.”
Also, they tried to analyse the offensive (16 percent of the content), highlighting the failure of the Russian military command to learn after the Ukrainian Kharkiv counteroffensive in 2022 and to supply the army with the necessary equipment.
A peculiar distinct topic was about New York – a settlement on the frontline in Ukraine. The war correspondents were cheering up the audience, citing the coverage by a journalist Christopher Miller that featured a phrase “New York will be captured, 100 percent.” Additionally, war correspondents mentioned the Russian offensive in the Donetsk region: “Much worse brigades appeared near Toretsk with the aim of improving the position of the Ukrainian Armed Forces. However, for some reason the Ukrainians struck in the main direction of the Kursk NPP. Probably, they decided to repeat the raid tactics of the Israeli Moshe Dayan”
Also, they mentioned the affected civilian population in 11 percent of the content, making it the fourth most popular topic. There were two unique topics in war correspondents’ content that we have not identified across other sources. One was the issue of Orthodox churches and monasteries under fire due to the active combat operations (6 percent of the content). Another consisted of calls to donate to Russian military units defending the Kursk region (5 percent).
General population
The findings in this chapter are based on the comments of Russians in different social media with detailed exploration of the comments in Telegram and VK. Additionally, we featured findings from YouTube comments. In total, we gathered 160,670 comments.
Apart from identifying discussions of the combat operations (52 percent of the content) we have found a particularly interesting narrative that highlights the sceptical approach of the Russian citizens towards the government – 12 percent of the comments in Telegram were about the inability of the army to prevent the incursion.
A portion of the discussion (6 percent) consisted of concerns about the Ukrainian army advancing towards the Kursk NPP and potential occupation of the plant. Moreover, the “nuclear discourse” featured calls to use the tactical nuclear weapon against Ukraine and Western countries – such a narrative featured in 3% of the comments.
Almost 6 percent of the comments were related to grievances about the lack of social support (such as compensations for lost assets, monetary help for the affected population) of residents of the affected areas: “It’s impossible to receive compensation for loss, bombed housing, car, the authorities have huge number of complaints”. These discussions contrast with the pro-Kremlin media that made the help for locals a major topic in their broadcast: “The authorities present some minor assistance as something incredible, their achievements. It contradicts common sense. The authorities, this is your direct duty”. Also, people connected the lack of monetary help to high spendings on war effort: “Putin promised to pay 10 thousand rubles [~110 USD – auth.] to residents of the Kursk region who suffered from the war. Russia is a generous soul. Sorry, there is no more money, but you hold on. Everything went to the war and war correspondents”. Moreover, not only Kursk region residents experienced issued – they applied to the inhabitants of Belgorod region, too: “Several thousand residents of Shebekino, Grayvoron and other districts cannot leave their homes, there is nowhere to go, there is a queue at the temporary accommodation centre”
The unique topic featured in this section is the question of civilian cars usage in combat areas. Commenters raised concerns that actual civilians from the Kursk region driving the cars might be considered as combatants since both Ukrainian and Russian armies actively use civilian transport for their own needs.
We have made a deeper exploration of the 30 percent comments that analyze combat operations and 12 percent that blame Russian authorities to highlight grievances and discontent. Within this topic, 42 percent of comments consisted mostly of commenters being angry at the Ukrainian army featuring derogatory language towards it and Ukraine in general. However, another 58 percent contained peculiar narratives.
The most popular one was about the Russian authorities and army being ineffective in preventing the incursion and managing the defence. Another 11 percent consisted of conspiracies that Russian military command consists of state traitors, mostly mentioning the Chief of the General Staff Valery Gerasimov in this regard. This is echoed by YouTube comments, 6 percent of which consist of blaming Gerasimov for failures to defend the Kursk region.
Smaller topics featured blaming local municipalities for not evacuating the civilians before the incursion (7 percent), blaming authorities for not letting the local population to use their weapons against the Ukrainian army (6 percent), and the despair with “Akhmat” special forces unit that failed to effectively defend the region (4 percent).
Below are some comments featuring disassociation of official narratives and the perception by the people:
“This is a lie. The situation is as follows: the entire settlement of Sudzha, the village of Kromskiye, Byk, and the Lgovsky district are controlled by the Ukrainian Armed Forces. The latter have evacuated. They say that in Malaya Loknya, corpses are lying right on the road, there are mines everywhere, and Ukrainian Armed Forces sabotage groups are advancing. In Malaya Loknya, according to the latest information, there is a checkpoint with a machine gunner at the exit from the village”
“Lately they've been talking nonsense. Perhaps there are no guidelines in connection with the recent events in Kursk”
“The territories occupied by the Ukrainian Armed Forces in Kursk are comparable to those captured with enormous efforts in about a year by the Russian Army”
Additionally, the uncertainty and lack of confidence in the Russian army and the authorities featured disbelief in the “red lines” imposed by the Russian political establishment, and despair in military effectiveness of the army:
“It is absolutely correct to call what is happening this way: the front has collapsed, and there is a complete failure in the northern direction.”
“There is nothing to defend with, everything has been stolen. That's for sure. The red line has already been crossed.”
“It is very sad that in recent weeks some people have been trying to force false panic moods, especially regarding the Kursk Nuclear Power Plant. Some are already suggesting fleeing to the Urals or even drowning themselves in the Pacific Ocean."”
“Kursk under fire!!! Where is Akhmat? Where is Wagner? Where is the army? Where to go? What to do?”
In VK, the most popular topic consists of more than a quarter of comments and is related to the corruption in the Russian Ministry of Defense – a condition that made the incursion real, according to the commenters. Within these grievances the most notable were referring to the foreign agents, the deferment from military service, and the need to continue changing personnel in the Ministry. Another peculiar topic included ironising about the Russian army instantly retreating to Moscow instead of fighting along the border.
Also, a notable narrative is the discussion of the fate of the prisoners of war captured by the Ukrainian army since there were videos with hundreds of Russian soldiers surrendering – this narrative made up 11 percent of the VK comments. Within the 11 percent, a fifth of discussions contained blaming the military command for ineffective rule. Additionally, a small portion (1 percent of the comments) featured the discussions about conscripted soldiers taking part in combat operations among other grievances about the incursion:
“If "there are enough fighters at the front", why were conscripts with small arms guarding the state border in the Kursk region, right next to the front?”
“Why write such things about Kursk, civilians are dying there, conscripts are being taken as POWs”
“Where is the overhyped multi-layered defence? Where have the fighting Chechens disappeared? Why are there only conscripts and border guards in the path of the Ukrainian army?”
Notable topic is appealing to PMC Wagner as a potential actor in defending the Kursk region (7 percent), pointing out their effectiveness compared to regular units of the Russian army.
Another 7 percent of the comments not featured on the chart are minor topics like wishing luck to Russian soldiers and praying for them.
Bots
We have identified that 2.7 percent of the comments from general population came from bots. The most popular topic included describing the Ukrainian army as war criminals – 18 percent of the content. Within this topic bots stressed out the usage of chemical weapons on the frontline by Ukraine and actively exploited the topic of affected civilians, stating that Ukraine is responsible for the “terrorism against the civilians. Other topics included exaggerating the losses of the Ukrainian army (14%) and criticising the political leadership of Ukraine (14%).
Regional outlets
Official authorities
In total, we gathered 994 publications from 15 official channels. Local officials in Kursk (district administration and particular political figures like the head of the Kursk region) mostly stressed security of the civilians: almost half of the content was related to bomb shelters, evacuation, and temporary residence of the affected population. The second topic featured reporting about the shellings aftermath, such as results of work by Russian anti-aircraft weapons or affected buildings in the region. The third most popular topic (25%) consists of formal notifications about air raids from the Ukrainian army.
Local groups
In total, we gathered 4,475 documents from 32 local channels.
Among regional communities in Telegram, the most popular topic, making up 35 percent of the content, was reporting on combat operations in the region – citing Russian Ministry of Defence briefings or war correspondents on the spot. The second topic (34 percent) was related to humanitarian aid for the locals and informing about potential evacuation measures. Also, it featured content about bomb shelters in the region and guides on safety during shelling – such publications contain advice related to physical security: “If you are at home, seek shelter in a room with windows and solid walls, such as a hallway, bathroom, or closet. If you are outside, go into the nearest building or suitable shelter.” Similar topics were echoed by the commentators in such communities.
The third topic (31 percent) was the air threats from the Ukrainian army and mostly consisted of notifications about air alerts in the region. Also, this topic featured advice of how to behave in a combat zone: “Dear Kursk region residents, do not film or post online the work of the air defence systems, the movement of our troops, and do not engage in dialogue with strangers on social networks who are trying to find out the situation and the presence of troops in your area.”
Among popular local groups in VK the most popular topic was related to civilians, making up 30 percent of the content. It consists of the posts about the evacuation, the effect of combat operations on civilians and reporting on medical aid for them, with posts of casualties among local residents in the area of the incursion.
The second topic, making up 27 percent of the sample, is related to analysing the motives behind the incursion. According to this analysis, the key reasons might be to complicate the logistics of the Russian army, to make the Russian army rotate their units from the east of Ukraine, to stop the gas supply from Russia to Europe, and to improve the position of Ukraine before the potential peace negotiations: “Ukraine invaded the border districts of the Kursk region of the Russian Federation to improve its negotiating position. The positive opinion about the negotiations is due to the fact that Russia's loss of people, equipment and territory, as well as the advance of the Ukrainian Armed Forces deep into Russian territory, should frighten the Russians.” Also, this topic featured analysis of the goals of the Ukrainian offensive: “A realistic plan for the Ukrainian Armed Forces could include a maximum advance in three or four directions deep into Russian territory, the occupation of large historically famous cities and regional centres such as Rylsk and Lgov. This would expand the line of combat contact and complicate the supply of the Russian Armed Forces groups.”. Another point of view included assuming that the whole incursion is designed only for boosting morale and not for any military goals: “The operation may simply be a PR stunt to maintain the morale of the Ukrainian military and civilians, a test of strength and so on.”
The third biggest topic (26 percent) was dedicated to reporting the combat operations, like the following example: “the enemy launched an attack on our positions in the southeastern part of Sudzha. Serious firefights are underway, the enemy's concentration remains.”
Five percent of the publications featured alarms about potential combat operations near the Kursk NPP and 2 percent was about blaming NATO and the West for an incursion.
Among the small but peculiar narratives were concerns about the potential mobilisation due to the incursion and appealing to the international law, saying that the Ukrainian army violated it – each of the topics made up 1 percent of the content.
Conclusions
The recent incursion of the Ukrainian army into Russia's Kursk region has created an effect of significant anxiety and doubt across social media – a similar situation to mobilisation in September 2022 and Evgeny Prigozhin's mutiny in June 2023. While Russian authorities attempt to portray the Ukrainian operations in Kursk as a waste of resources without any military goals, the response from the general population highlights a concern about the state's ability to protect its own territory and citizens. Although the overall war support significantly dropped during the first week of the incursion due to the unexpected nature of the event, the long-term effect remains uncertain yet. This is tied to the local nature of the operation (the fighting takes place in several districts of the Kursk region). Therefore, it is early to make global conclusions about the war support from Russian citizens. However, we will explore further developments of this situation in subsequent reports.
Reprinted from https://www.openminds.ltd/ with the author's permission. You can find the original article here.
OpenMinds develops technologies and tools to fight propaganda and disinformation. Since 2022, OpenMinds has participated in 240+ counter-propaganda operations and published reports featured in The Economist, Times, Guardian, New York Times, CNN, and Financial Times.
You can also highlight the text and press Ctrl + Enter
