by Amelia Arsenault and Sarah Kreps
Arsenault, A. C., & Kreps, S. (2026). Whose voice counts? The role of large language models in public commenting. Big Data & Society, 13(1). https://doi.org/10.1177/20539517261419341 (Original work published 2026)
Large language models (LLMs) may help bridge the accessibility gap in public commenting by helping citizens navigate dense policy documents and draft effective comments. At the same time, they may also exacerbate existing inequalities. To examine this tension, we conducted a survey experiment in which participants drafted public comments with and without LLM assistance. We find that LLMs made writing significantly easier across all education levels but did not improve policy comprehension. Instead, education remained the strongest predictor of understanding. Comments written with AI assistance received consistently higher quality ratings regardless of education level, suggesting that LLMs may help less-educated citizens meet the threshold for serious consideration by policymakers. Although these tools may expand access for historically underrepresented voices, broader investments in civic education and AI literacy remain essential for achieving substantively meaningful democratic engagement.
