A new international study revealed that artificial intelligence responds most effectively when prompted in Polish, while English only ranked sixth. Researchers from the University of Maryland and Microsoft examined 26 languages to identify which ones AI systems process with the highest accuracy.
The research team evaluated several leading AI platforms—OpenAI, Google Gemini, Qwen, Llama, and DeepSeek—using identical inputs in each language. Their analysis overturned assumptions about English dominance in AI communication.
“Our tests delivered results few expected,” the report stated. “Polish ranked first for comprehension and performance, whereas English finished sixth when handling long and complex prompts.”
Machines Excel at a Language Humans Struggle With
According to the study, Polish achieved an average accuracy of 88% in executing commands. “Polish turned out to be the most precise for interacting with AI,” wrote the Polish Patent Office. “It challenges the belief that Polish is too hard—humans find it difficult, but AI does not.”
Experts noted the irony that AI mastered Polish despite limited training data in the language compared to English or Chinese. Surprisingly, Chinese placed near the bottom, ranking 23rd out of 26 languages tested.
How the Languages Ranked in AI Efficiency
The top-performing languages for AI understanding were:
- Polish — 88%
- French — 87%
- Italian — 86%
- Spanish — 85%
- Russian — 84%
- English — 83.9%
- Ukrainian — 83.5%
- Portuguese — 82%
- German — 81%
- Dutch — 80%
The findings suggest that AI may not favor the world’s most widely spoken or simplest languages. Instead, structure and syntax could play a greater role in how machines interpret human communication.
