Thread by @GrandpaRoy2

Russian blogger Alexei Rogozin warns that the American company Palantir has provided Ukraine with AI tools that process vast amounts of data to integrate with and leverage long-range UAV strike capabilities.
1/

[Tuapse oil terminal burns]

https://twitter.com/grandparoy2/status/1904985874712150383

“Ukraine's long-range strikes on Russian targets cannot be viewed solely as a problem of air defense. An unmanned aerial vehicle (UAV) or missile is just the final link in the chain.
2/

“The main work begins earlier: reconnaissance, accumulation of digital traces, analysis of satellite and aerial photos, comparison of open and closed data, assessment of the vulnerability of objects, and review of the results of previous strikes.
3/

“The danger lies not only in the UAV itself, but in the entire system that helps to understand in advance where, when, and with what calculation to send it. Ukraine's 35-year-old Defense Minister, Mikhail Fedorov, openly stated that cooperation with the…
4/

“… American corporation Palantir has given Ukraine tools for analyzing aerial strikes, artificial intelligence solutions for processing large amounts of intelligence data, and the integration of these technologies into long-range strike planning.
5/

[FP-1 UAVs launching]

“The point is not that the American program itself chooses the target. The purpose of such systems is different: they allow to quickly collect fragmented information into a single picture.
6/

“Satellite images, UAV videos, surveillance data, interceptions, information about the repair and restoration of objects, repetitive routes, the results of previous attacks.
7/

“For long-range strikes, it is crucial to understand what function the object performs, how it is related to production, logistics, or energy, how quickly it is restored, and what effect a repeated hit will have.
8/

“Palantir here acts as a data processing and linking environment, while the Ukrainian system Delta provides a unified picture of the situation, linking UAVs, sensors, units, and strike means.
9/

[Fedorov and Zelensky meet with Palantir CEO Alex Karp]

“Ukraine has already been able to integrate drones, sensors, and strike means into a common network faster than the US Army itself.

Also indicative is the Ukrainian project Brave1 Dataroom, created together with Palantir.
10/

“It is positioned as a closed environment for training artificial intelligence models on real combat materials. More than 100 companies are officially involved and over 80 models are being trained, primarily for the detection and interception of targets like ‘Geran’.
11/

“In fact, this is an attempt to turn war into a continuous learning cycle: collect data, label it, train the AI model, apply it, measure the result, and improve the system again.
12/

[Russian ‘Geran-2’ UAV is destroyed by ‘Sting’ interceptor]

“From this follows a simple conclusion: the fight only against unmanned carriers does not solve the problem. If the enemy constantly updates data about the object, understands its role in the industrial or military chain, …
13/

‘… analyzes the consequences of strikes, and quickly trains new models on combat experience, then the strike UAV itself becomes just an executing element.

Therefore, we need not only air defense and electronic warfare.
14/

We need camouflage, reduction of the observability of objects, control of digital traces, protection of logistics, false signs, rapid analysis of the consequences of strikes, and a unified data system for object defense. The enemy is not just building a fleet of drones, …
15/

“… but a full-fledged reconnaissance and strike architecture. The response must be of the same level. In modern warfare, the one who learns to quickly turn every combat episode into data, a decision, and ultimately a new quality of management will win.”
16/16

Sources
t.me/rogozin_alexey…
t.me/rogozin_alexey…