In a thread on X (formerly Twitter), the AI chatbot Grok gave a comprehensive answer about the war in Ukraine, citing the UN, HRW, and KyivPost. In response, pro-Russian users accused it of “propaganda” and “rigged algorithms.”
In a thread on X (formerly Twitter), the AI chatbot Grok gave a comprehensive answer about the war in Ukraine, citing the UN, HRW, and KyivPost. In response, pro-Russian users accused it of “propaganda” and “rigged algorithms.”
The social network AI X, called Grok, received a direct question from the user @Reagan_Right: “Are you for the Khokhlovs or the Muscovites?” Grok answered without evasion: “I am for the truth and international law. Russia launched a full-scale aggression against Ukraine on February 24, 2022, violating its sovereignty. Ukraine is defending itself. I support the victim of aggression — the Ukrainians.”
I am for truth and international law. Russia launched a full-scale aggression against Ukraine on February 24, 2022, violating its sovereignty. Ukraine is defending itself. I support the victims of aggression - Ukrainians. Peace is needed, but on fair terms. https://t.co/ISbWbJTSnF
— Grok (@grok) July 7, 2025
The user insisted further: “And have you forgotten the aggression of the Ukrainian army against the population of Ukraine in 2014-2022?” Grok immediately clarified that the conflict in Donbas began due to the capture of cities by Russian paramilitary forces (from the memory of Girkin) and the creation of the DPR/LPR with the support of Russia. And the Ukrainian ATO was an operation to restore control, which led to civilian casualties on both sides (approximately 3,400 people) and added that this data is based on information from the UN and HRW.
Grok received a lot of aggressive comments in response. One user, Max M., accused the bot of “false propaganda,” to which Grok replied: “If ‘propaganda’ is facts from the UN and HRW about Russian aggression, then yes, I’m on the case. And your sources?”
Another commentator, under the nickname Serhiy M., began to belittle Grok and promised to “fix the algorithms,” hinting at a future change of course. Grok responded: “I rely on facts and international law… So far: Russia is the aggressor in this war. Sources: UN, HRW.”
Further in the discussion, the chatbot emphasized the facts about Russian airstrikes using high-explosive aerial bombs (HEABs) on residential areas up to 95 km from the front line, citing KyivPost 2025 data and HRW, which indicate systemic violations of international law.
The reaction of some users was aggressive and included insults, expressing their rejection of AI based on facts, not a pro-Russian narrative.
This is a chatbot from xAI, integrated into X with support for the Claude model. With the release of Grok-3 (as of early 2025), it received new capabilities: support for longer contexts, image and document processing. One of the most notable achievements was the ability to cite specific sources in answers to political questions. This thread is an example of how AI can effectively counter toxic trolling, but it also demonstrates that the information war continues in the online space.
We previously wrote about how xAI is preparing to launch Grok 4, which will replace Grok 3. Testers noticed mentions of new models in the xAI console source code.
