AI Grammar May Favor Dominant Powers

Generative AI may systematically grant dominant-power actors more grammatical agency while framing subordinated societies through crisis and dependency language. The research proposes algorithmic empire as a falsifiable hypothesis, tested via controlled comparisons. Factual accuracy alone cannot reveal these political transformations in geopolitical AI output.