Language changes political views in AI answers on Ukraine war
Geopolitical Divisions Across Languages in Large Language Models
Artificial IntelligenceComputation and LanguageComputers and Society
Summary
The authors found that AI chatbots give different political answers about the Ukraine war depending on the language of the question. They tested three AI models in 112 languages and saw shifts between Russia-leaning and Ukraine-leaning responses. These language differences align with real-world political views and actions by countries. This means AI might reflect geopolitical biases based on the text it learned from.
What this means in practice
- •For multilingual ai developers: Detect and mitigate geopolitical bias in AI responses across different languages to improve model fairness and reliability.
- •For international news platforms: Use insights on language-driven bias in AI to provide more balanced multilingual explanations of global events.
Authors
Maxim Chupilkin
Abstract
People increasingly turn to AI chatbots for news and explanations of world events. But do they receive the same political answers when they ask in different languages? Here we show that the language of a question can change how the same AI systems assess the war in Ukraine. We ask GPT, Claude and Gemini to evaluate twenty statements about the war in 112 languages, collecting 67,200 responses. The balance between Russia-leaning and Ukraine-leaning responses differs across languages. When we group responses by countries' official languages, they follow a pattern resembling worldwide political divisions: relatively more Russia-leaning answers correspond to more favourable public views of Russia, less support for Ukraine in United Nations votes, and less aid to Ukraine. The broad pattern recurs across all three models and remains when individual statement pairs are removed. Our findings suggest a possible route through which information warfare may shape the text used to train AI models, which may in turn spread geopolitical biases.