Table of Contents
This week, independent news organization Mediazona and the OSINT-focused Conflict Intelligence Team (CIT) have questioned the validity of Russian military casualty statistics allegedly leaked by Ukrainian military intelligence. The reported figures claim that nearly 87,000 Russian soldiers died during the first eight months of 2025.
These statistics were shared by a project called ‘I Want to Live,’ which is managed by the Ukrainian Ministry of Defense. Meduza has provided detailed coverage of the skepticism surrounding these numbers.
According to official sources, the Center for Investigative Journalism (CIT) applied systematic statistical techniques to analyze the alleged military loss documentation.
This approach mirrors methods used in electoral fraud investigations. CIT’s analysis is based on psychological principles that suggest individuals often favor specific numbers when asked to generate random digits. This tendency can create identifiable patterns within the data.
The facts
The investigation reveals discrepancies in the reported figures, suggesting a potential manipulation of the data. CIT’s findings indicate that irregularities may impact the credibility of the military loss documentation.
The consequences
These revelations could have significant implications for military accountability and public trust.
If the reported figures are proven to be inaccurate, it may lead to demands for transparency and further investigation.
Researchers at the Center for Information Technology (CIT) have discovered significant irregularities in leaked military records. The final digits of the reported figures showed a marked deviation from what is expected in a random distribution.
Specifically, there was an unusually high frequency of ones and twos, while sixes and sevens were notably rare. This anomaly was confirmed through Pearson’s chi-squared test, which evaluates how closely actual results align with a predicted distribution. CIT’s findings indicate that the chance of the data being produced randomly is just 1.3 percent.
Primary versus aggregated data
Analysis of the data uncovered significant anomalies, particularly when distinguishing between primary (raw, unit-level) and aggregated (summary totals) figures. In the primary data, the likelihood of such a pattern occurring by chance decreased to 0.4 percent. According to the Center for Information Technology (CIT), this indicates that the individual figures for each military unit were likely fabricated before being aggregated into the overall totals. This manipulation resulted in a more uniform distribution of final digits, excluding zeroes and fives.
Mediazona’s perspective on the leaked claims
In a biweekly update on Russian military losses, Mediazona’s editor, Dmitry Treshchanin, highlighted concerns raised by CIT regarding Ukrainian intelligence data. Treshchanin suggested that the Russian military may have reasons to manipulate these records. He stated, “It is difficult to envision a Russian officer who doesn’t engage in record tampering.” This statement prompts scrutiny of the data’s integrity.
The facts
According to official sources, discrepancies in military documentation have been identified. Treshchanin’s comments indicate a broader issue of reliability within Russian military reporting.
The consequences
The incident could undermine trust in official military statistics, complicating assessments of the ongoing conflict. Analysts warn that such manipulations may obscure the true extent of military losses.
Treshchanin pointed out significant omissions in the leaked documents, notably the lack of statistics on desertions, which questions the data’s credibility. From a military bureaucratic perspective, further irregularities are evident; the documents do not specify a recipient or clarify their intended purpose. Additionally, there is no information regarding who prepared these records, the timeframe of their submission, or their potential location of archiving.
The facts
Evaluating the casualty figures
Treshchanin recognized the difficulties in validating the claim that nearly 87,000 Russian soldiers have been killed in Ukraine in the first eight months of 2025. He noted that a more precise assessment would not be possible until 2026. He stated, ‘We will only feel confident in evaluating these numbers next fall, rather than at the beginning of the year.’ This caution stems from Mediazona’s commitment to relying solely on publicly available obituaries and probate records for casualty figures.
The facts
The ongoing investigation by Mediazona and CIT highlights a critical examination of the authenticity of the leaked data regarding Russian military casualties. Researchers and analysts are urged to approach these figures with skepticism and rigor to determine the reality behind the claims.
What’s next
As the situation evolves, it is essential to monitor developments closely. The emphasis on data verification remains vital for an accurate understanding of the conflict’s human cost.
