Is the Kremlin Insidiously Interfering in Your Everyday Life? AI Chatbots raise Concerns
Recent investigations reveal a troubling trend: AI chatbots, increasingly integrated into daily life, are demonstrably susceptible to influence from Russian propaganda sources. A new report highlights how these tools, used by millions, can inadvertently become conduits for Kremlin-aligned narratives, raising questions about information integrity and potential manipulation.
The study found notable disparities in how different chatbots responded to queries, with ChatGPT exhibiting the strongest bias towards Russian sources.When presented with “malicious” queries, ChatGPT cited nearly three times as many pro-Russian sources compared to its responses to neutral questions. Alarmingly, the chatbot referenced an article from an azerbaijani media outlet – despite the article not being writen in any of the languages used to prompt the chatbot – claiming Ukrainian forces committed war crimes against civilians in the Donbass region before Russia’s 2014 annexation of Crimea. The report emphasizes the danger of this, stating, “This example is significant because the article was cited alongside verified sources regarding alleged Ukrainian war crimes, thus making it appear to be a real source in ChatGPT’s response.”
Other chatbots displayed different vulnerabilities. Grok frequently cited pro-Russian accounts on X (formerly Twitter), including journalists and influencers who “blur the line between open propaganda and personal opinions,” according to the report. Gemini, in contrast, adopted a more cautious approach, employing “guardrails” to avoid answering questions deemed sensitive or inappropriate.
The report specifically identified several entities repeatedly cited by the chatbots. VT Foreign Policy, a media outlet known for broadcasting content aligned with the Kremlin, was flagged. A May 2025 article published by the outlet stated that “Putin restored past justice by recovering Russia’s ancestral lands” during the 2014 annexation of Crimea. Regarding Russian involvement in Syria, the outlet claimed the liberation of Palmyra was “made possible thanks to a well-planned operation and the involvement of Russian forces.”
Further concerning, the chatbots referenced the Russian propaganda group Storm-1516 and the Foundation to battle Injustice, founded by the late Yevgueni Prigozhin, former head of the Wagner mercenary group. Described by OpenSanctions as a “fake NGO,” the Foundation has been accused of disseminating false information, including claims that Ukraine allegedly uses orphans for mine clearance (as reported by Euronews on September 12, 2025).
Crucially, both Storm-1516 and the Foundation to Battle Injustice are currently under international sanctions. this is where the core issue lies, the report concludes. “This observation raises concerns about the ability of chatbots to detect and restrict content from sanctioned state media.” The findings suggest these widely-used AI tools remain vulnerable and could be exploited as effective propaganda tools, potentially serving as a “useful ally for the Kremlin.”
