When “Government-Approved” LLMs Silence Citizen Voices: Evidence of Systematic Demographic Bias in Federal Comment Analysis

Sola Kim

Keywords: large language models, fairness auditing, bias evaluation, interpretability, social aspects of machine learning

Submission Number: 356