The quiet way AI normalizes foreign influence

This analysis argues that large language models, constrained by source availability and paywalls, disproportionately cite state-aligned media, steering users toward propaganda; FDD research found 57% of responses about conflicts cited such sources and 70% of neutral Israel–Gaza queries cited Al Jazeera. It warns that AI citation practices shape public trust and traffic, potentially disadvantaging independent journalism while amplifying foreign influence operations. The piece urges licensing of credible journalism, labeling/deprioritizing state-controlled outlets, and integrating citation patterns into AI safety frameworks, aligning with recent White House guidance on ideologically neutral AI procurement.