A recent investigation has identified a growing trend of AI-generated content infiltrating the Australian parliamentary inquiry system. These inquiries, designed to gather expert testimony and public feedback to inform legislation, are reportedly being flooded with submissions created by large language models (LLMs).

According to reports, these AI tools are producing "hallucinations"—content that appears credible but is entirely fabricated. The issue is particularly concerning as these submissions often invent academic studies and attribute nonexistent research to real-world experts and authors. In some instances, committee reports have inadvertently cited these AI-generated submissions as legitimate evidence.

Experts warn that if this trend persists, it could undermine the reliability of the parliamentary process. Because the inquiry system relies on the assumption that submissions are authentic contributions from citizens and professionals, the introduction of automated misinformation complicates the ability of lawmakers to make evidence-based decisions. While the full scope of the infiltration is still being assessed, the presence of these hallucinations suggests a significant challenge for parliamentary staff who must now verify the authenticity of an increasing volume of digital submissions.