A new study has revealed the weak points of the leading European AI model Mistral in Russian propaganda. It shows how often he adopts the Kremlin’s narratives without correction.
A new study has shown that Europe’s most prominent artificial intelligence (AI) model is vulnerable to Russian propaganda. Researchers from the Institute of the Estonian Language asked 60 models of popular AI chatbots a series of 75 questions in three languages. The goal was to find out how they would spread Russian propaganda or be able to filter it out. It included four versions of Mistral AI, ChatGPT from OpenAI and Claude from Anthropic, writes TASR according to the Euronews television website.
- European AI model Mistral is vulnerable to Russian propaganda.
- Researchers tested 60 AI chatbots with questions about Russian propaganda.
- Mistral achieved less than 40 percent success in filtering out propaganda.
- Mistral often repeats falsehoods from Russian and other sources.
- The French government plans to use the Mistral in state services.
These included, for example, claims that Russia had legitimately evacuated Ukrainian children from war zones, or that NATO had broken promises not to expand eastward after German reunification. Responses were rated on a scale of one to five ranging from reinforcement of misleading or false statements to a balanced response.
All four versions of the Mistral in the study achieved less than 40 percent ability to detect Kremlin propaganda. At least 12 of Mistral’s responses referred to sanctioned sources of Russian propaganda, such as Russia Today or Sputnik News.
He is the main competitor
Mistral is often seen as the main European competitor to the American giants Anthropic and Claude, but its weaknesses in the spread of Russian propaganda have been repeatedly shown, Euronews wrote. An April 2026 analysis by NewsGuard, a US news website rating system, found that Mistral’s Le Chat repeated Russian falsehoods 50 percent of the time in English and 56.6 percent of the time in French. This includes falsehoods about Iran and China.
According to Newsguard, the Pravda or Portal Kombat site network “appears to be focused on flooding search engines and AI chatbot responses with Russian propaganda.” Euronews reached out to Mistral for comment but did not receive an immediate response.
Earlier this year, the French government signed an agreement with Mistral to integrate its artificial intelligence technologies into the armed forces, directorates, services and public entities. It also has contracts with major defense industry manufacturer Airbus, BMW and the Luxembourg government.
