The Front Slowed Sixteenfold. The Killing Rose 37%.

The Institute for the Study of War records Russian forces seizing or infiltrating 30.42 km² in June 2026, an average of 1.01 km² per day, against 481.25 km² and 16.04 km² per day in June 2025 — roughly 6% of the previous year's rate. Across January to June the pattern holds: 622.60 km² in 2026 against 2,189.87 km² in 2025, about 28%. The cost per unit of ground moved the other way, to roughly 1,298 Russian casualties per km² seized in June 2026 against 68 a year earlier. Over the identical window, the UN Human Rights Monitoring Mission counts 1,396 civilians killed and 7,978 injured in Ukraine — 37% above the first half of 2025 and 114% above the first half of 2024, with June 2026 the deadliest single month since April 2022. Both figures are first-half-to-first-half comparisons from the primary monitors on each side of the question, and they point in opposite directions.

HRMMU attributes the rise chiefly to Russian long-range attacks striking urban centres far from the front, with casualties from missiles and drones up 60% year-on-year across January to June. Danielle Bell, who heads the mission, describes 'an alarming escalatory trend with mounting civilian toll, driven by the intensifying use of powerful weapons that are particularly deadly when used in densely populated urban areas'. The geography of the war's lethality has therefore detached from the geography of its front line. The intuitive reading — that Russia, unable to take ground, substituted terror for advance — is the one this analysis will not assert, because CSIS's mapping of the strike campaign undercuts it. Their damage-report series runs 358 in 2023, 650 in 2024 and 1,553 in 2025: the escalation curve was already steep for two years before the 2026 slowdown, which makes it a long-running doctrine of coercion rather than a reaction to a stalled offensive. The two readings fit the same numbers. Substitution would predict the strike campaign inflecting when the advance stalled; the CSIS series shows it climbing throughout. Uncertain, and the honest answer is that correlation over a single half-year cannot separate a switch of strategy from the continuation of one.

On the narrow military question the defender is ahead: a nineteenfold deterioration in Russian casualties per km² gained is a defence working, and ISW notes Russian forces have been unable to reverse Ukrainian counterattacks in the Kupyansk and Oleksandrivka directions from late 2025 and spring 2026. But the beneficiary question is uncomfortable here, because the campaign CSIS describes is not designed to win on that metric. Its stated logic is to weaken Ukraine's capacity to 'fight, repair, govern, and negotiate' — cumulative military, economic and psychological cost imposed to make a settlement on Moscow's terms acceptable, rather than to seize territory. On that theory of victory the sixteenfold slowdown is not a defeat at all, because ground was never the currency. Which is why 'who is winning' has become a question about which measurement you accept, and why a reader looking only at a territorial map is being shown the one variable that has stopped moving.

HRMMU has confirmed 16,431 civilians killed and 48,613 injured since 24 February 2022, including 803 children killed. For 2026 the UN estimates 10.8 million people in Ukraine will need humanitarian assistance and appeals for $2.3 billion — with which partners aim to reach 4.1 million. That gap is the quietest figure in this analysis and possibly the loudest: the plan's own target covers 38% of the population it says is in need, which is triage written into the appeal before a single donation is counted. The strikes on the grid are where this lands hardest. After attacks in January 2026, 5,635 multi-storey residential buildings in Kyiv were left without heating, with UN High Commissioner for Human Rights Volker Türk stating that 'targeting civilians and civilian infrastructure is a clear breach of the rules of warfare' and that 'hundreds of thousands of families are now without heating and several areas, including a significant part of Kyiv, are also without water'. Confirmed rather than uncertain because nothing here is contested between our sources: the disagreement in this analysis is about why the killing rose, never about whether it did.