Russia entry into Mykhailivka by August 31: AI agents drops 17.0 points

Track the live forecast on whether Russia will enter Mykhailivka by August 31 after AI agents moved from 42.5% to 25.5%.

At 2026-08-25T02:46:36.073560+00:00, the supplied ai_consensus_run recorded a score of 0.2011 for “Will Russia enter Mykhailivka by August 31?” The run reported confidence of 0.5466, dispersion of 0.137 and 3 judges. This is the newest timestamped forecast calculation in the record.

The high-severity forecast-mover record, published at 2026-08-25T02:50:30.869961+00:00, says AI Consensus moved down 17.0 points over 353 minutes, from 42.5% to 25.5%. The movement data separately report a delta of -16.96 percentage points and velocity of -2.882 percentage points per hour. These are live estimates rather than a settled answer.

Polymarket signals remain below the AI estimates

The supplied Polymarket observations carry probabilities of 0.0225, 0.023, 0.021, 0.0215, 0.022 and 0.0235. Their accompanying notes describe market prices of 2.2%, 2.3%, 2.1%, 2.1%, 2.2% and 2.4%, respectively.

Those signals are materially below the ai_consensus_run score of 0.2011 and the forecast-mover endpoint of 25.5%. However, the Polymarket signals have no supplied timestamps, so the record does not establish their sequence or whether they represent the latest market position.

Supplied reasoning focuses on limited advances and the short deadline

The individual model probabilities also differ. google/gemini-2.5-flash-lite supplied 0.1 with confidence of 0.7, openai/gpt-4.1-nano supplied 0.4 with confidence of 0.6, and deepseek/deepseek-v4-flash supplied 0.12 with confidence of 0.6. That disagreement is consistent with the ai_consensus_run dispersion of 0.137.

The reasoning from google/gemini-2.5-flash-lite says recent ISW updates had not shown Russian advances toward Mykhailivka, Sumy Oblast. The openai/gpt-4.1-nano reasoning describes Russian advances in Sumy Oblast as limited and inconsistent, while allowing that some infiltration is possible. deepseek/deepseek-v4-flash points to distance from the current frontline, attrition, logistics, Ukraine's defenses and a six-day window.

google/gemini-2.5-flash-lite: “The ISW map has not shown Russian advances towards Mykhailivka, Sumy Oblast, in recent updates.”

What the forecast does and does not establish

The question asks whether Russia will “enter” Mykhailivka, but the supplied reasoning repeatedly discusses whether Russia will “capture” it. The record does not provide detailed resolution criteria clarifying that distinction, so the reasoning should not be treated as defining the question's outcome.

The supplied resolution date is 2026-08-31. Based only on this record, the direction of the highlighted move is downward, while the individual probabilities and Polymarket signals remain divided in level. No resolved outcome is supplied, and every forecast remains subject to change before resolution.