A new study has found that popular artificial intelligence chatbots, including ChatGPT and Google’s Gemini, sometimes repeated narratives linked to Russian propaganda when responding to questions about the war in Ukraine.
Researchers said the systems occasionally relied on content originating from websites connected to pro Kremlin information networks, particularly when answering politically sensitive questions. The report warns that these sources can influence AI generated responses if they become part of the information available online.
The findings do not suggest that the chatbots were intentionally designed to spread propaganda. Instead, researchers say the issue highlights how large language models can be affected by misleading or coordinated online content, making it more difficult to distinguish reliable information from disinformation.
The study recommends stronger safeguards to limit the influence of manipulated online sources and calls for greater transparency about how AI systems generate answers on sensitive topics. It also urges users to verify important information with trusted sources, especially when seeking news about international conflicts or politics.
The report adds to the growing debate over the role of artificial intelligence in the spread of misinformation and the responsibility of technology companies to improve the accuracy and reliability of AI generated content.
