A March 2025 audit by NewsGuard found that major AI chatbots repeated false narratives from the Moscow-linked Pravda network in roughly one out of three responses. The network operates across approximately 150 websites, publishes in 46 languages, and generates about 18,000 articles per false claim. The AI providers caught in this audit include OpenAI, Google, Microsoft, and Meta. When researchers specifically tested chatbots on Nordic-language queries, pro-Russian disinformation appeared in up to 50% of responses.
Source: Read the original article
