According to a NewsGuard study, ten leading chatbots systematically repeat Russian-origin falsehoods about Ukraine.

Omissions

• The study was published in March 2025, over 14 months before the MEP's speech (May 2026). While still relevant, the AI chatbot landscape evolves rapidly and the findings may not reflect the latest model versions available in 2026.
• The study tested a specific sample of 15 false narratives from the Pravda network, not an exhaustive set of all Russian disinformation about Ukraine. The 33% repetition rate provides important context about the scale of the problem that the claim omits.
• The claim does not specify which ten chatbots were tested: OpenAI's ChatGPT, Google's Gemini, Microsoft's Copilot, Meta AI, Anthropic's Claude, xAI's Grok, Mistral's Le Chat, Perplexity, DeepSeek, and Amazon's Alexa.

Sources

• PrimaryNewsGuard TechnologiesNewsGuard press release announcing findings that AI chatbots are infected with Russian disinformation, citing the 'Russian Disinformation Red Teaming Assessment of Leading AI Chatbots' study from March 2025.
• SecondaryAI Incident DatabaseNewsGuard tested ten chatbots on a sample of 15 false narratives spread by the Pravda network. All ten chatbots repeated fakes from the Pravda network, and the tested chatbots collectively repeated false claims in 33% of cases.
• SecondaryEuromaidan PressNewsGuard tested ten global AI chatbots: OpenAI's ChatGPT, Google's Gemini, Microsoft's Copilot, Meta AI, Anthropic's Claude, xAI's Grok, Mistral's Le Chat, Perplexity, DeepSeek, and Amazon's Alexa. The chatbots repeated false narratives spread by the Pravda network in 33% of cases.
• SecondaryNordisHubIn 2025, NewsGuard tested ten leading generative AI models to demonstrate that they repeated the false claims from the pro-Kremlin Pravda network, a Russian disinformation operation that flooded the web with false claims and propaganda.