A new study suggests true AI neutrality is impossible. By analyzing how models describe political figures, scientists discovered that these systems align with the cultural and political values of their developers, often mirroring the worldview of their origin.