The Army That Eats Its Own

Wednesday 25 February 2026

Four years into Russia’s full-scale invasion, a grim secondary story has begun to harden into something closer to a pattern: growing reports that Russian soldiers are being brutalised not only by Ukraine’s firepower but by their own chain of command. These accounts vary in provenance and reliability, and none should be consumed with the credulous hunger of the internet. Yet taken together—across Russian investigative reporting in exile, Ukrainian and Western media coverage and the accumulating testimonies of families and servicemen—they sketch a coercive military culture in which fear is not a side effect of combat but an instrument of command. 

The vocabulary itself has become revealing. “Meat assaults” is a soldier’s phrase, not a doctrinal term—an ugly shorthand for repeated infantry attacks that treat human bodies as a consumable resource. Reuters, summarising Western assessments, has described Russia’s growing reliance on waves of small infantry assaults pushed into Ukrainian lines—a tactic that can gain metres, sometimes, but spends men at a rate that would have appalled even the Soviet General Staff. 

Alongside that tactical picture sits the disciplinary one. As early as 2023 the Guardian reported claims by Russian assault troops that so-called blocking or “barrier” units were placed behind them to prevent retreat, with threats of lethal punishment for those who tried to pull back. Later investigations—most notably a 2025 report described by the Guardian as drawing on testimonies, leaked material and complaint records—allege that violence against refusers has moved from rumour into something closer to an informal system, with some commanders accused of killing soldiers who resisted orders or tried to avoid high-risk missions. 

Even if one sets aside the most extreme allegations, the logic is recognisable to any student of authoritarian militaries. When a campaign is politically non-negotiable but militarily expensive, the institution tends to shift risk downwards. Officers protect themselves through paperwork, intimidation and the ruthless allocation of danger. Soldiers become, in effect, the shock absorbers of strategic failure. Russia’s penal formations and “assault” detachments—an evolving category since 2023—have only sharpened this logic, because they normalise the idea that some lives are simply cheaper than others. 

This matters for a second reason: it may distort our understanding of casualties, including the ratio between Russian and Ukrainian losses at the front.

Casualty counting in this war is a foggy business. Neither Moscow nor Kyiv offers transparent, independently verifiable figures, and in any case “casualties” often means killed, wounded and missing combined—useful for gauging military strain but easy to misunderstand. Reuters reported NATO’s claim of around 65,000 Russian casualties over the last two months and referenced CSIS estimates of roughly 1.2 million Russian casualties since 2022, while noting the Kremlin’s denial and the practical difficulty of verification. CSIS, in its own published analysis, argues that Russia’s losses are driven in part by a strategy of attrition—wearing Ukraine down by repeatedly feeding manpower into the fight.  Mediazona’s work on verified deaths and Russia’s probate registry provides a different methodological lens, explicitly designed to compensate for propaganda and silence, and it continues to update its lists and estimates. 

So where does internal mistreatment fit?

First, coercion makes waste easier. A commander who can compel assaults through fear can keep attacking even when battlefield logic says to pause, reconnoitre and husband forces. The tactical result is a higher speed of infantry pushes—more exposure to drones, artillery, mines and small arms—hence more Russian losses for small territorial returns. Reuters’ own framing is blunt: modest gains, steep cost, sustained by repeated infantry assaults. 

Secondly, coercion changes what “avoidable” means. In a normal army, a soldier who is exhausted, injured or traumatised still generates institutional incentives for treatment, recovery and reconstitution. In an army where intimidation and informal violence are reported as tools of discipline, the soldier becomes less a trained asset than a replaceable body. That attitude is consistent with the recurring Western and Ukrainian descriptions of Russia’s manpower model: recruitment, short training, then forward. It is a model that can keep a front from collapsing, but only by chewing through people.

Thirdly, it may skew the casualty ratio further against Russia than earlier, more conservative estimates suggested. For much of the war many analysts assumed Russia’s losses were higher but not fantastically so—perhaps in the region of two Russian casualties for every Ukrainian casualty in equivalent categories, depending on period and sector. Yet by late 2025 the Economist—aggregating a large body of estimates—argued that, even under assumptions that would inflate Ukrainian losses, the ratio of Russian to Ukrainian deaths could be closer to five to one in certain frames of analysis. The point is not that any single ratio is “true” across the whole front. It is that the combination of tactics (repetitive assaults) and discipline (reported coercion, including lethal punishment for refusal) pushes the system towards higher Russian loss rates—because it reduces the practical ability of soldiers and junior leaders to say “no” to the battlefield when the battlefield is screaming “stop”.

One should add a final caution, because speculation is cheap. Ratios change by month, by sector, by mission, by weather and by the quality of fortifications. Ukraine’s defensive success often comes at a cost, particularly when she is short of artillery ammunition or forced to hold exposed positions. Meanwhile Russia’s sheer size allows her to conceal losses politically for longer than many outsiders would expect. Even the best open-source work is still an estimate, and the war is designed, by both sides, to make outsiders misperceive it.

But there is a sober inference that does not require sensationalism. An army that relies on fear to move forward will move forward more often than it should. An army that treats men as consumables will consume them. And an army that is prepared, even occasionally, to turn its weapons inwards is an army signalling that it has found a way to continue the war without solving the underlying problem of competence. In that sense, the reported mistreatment of Russian soldiers is not only a moral disgrace—though it is that—it is also a strategic indicator. It suggests that Russia is sustaining her war not through excellence but through coercion, and that the true human arithmetic of her invasion may be grimmer than the already appalling headline figures.