‘Scary’: how misinformation and AI hallucinations are infiltrating Australia’s parliament
Summary
Australia’s parliament inquiry system is receiving many submissions that include AI-generated false information, known as AI hallucinations. These hallucinations are false content made by large language models (LLMs), and some official reports have mistakenly cited this fake research as real. This raises concerns that lawmakers might base decisions on incorrect evidence.Key Facts
- AI hallucinations occur when AI like ChatGPT creates believable but false information.
- These AI errors happen because the AI predicts text but cannot tell if it is true or false.
- Some parliamentary committee reports have cited fake studies and incorrect research findings from AI-generated submissions.
- Google’s AI search summaries can also create fake references and cite those as real sources, worsening misinformation.
- An inquiry submission falsely attributed a study to a real academic, Divna Haslam, causing concern about misinformation.
- The organization that submitted the incorrect report said they used AI and tried to correct errors, but it was too late to update the submission.
- Committee members say they review all evidence carefully but receive materials of mixed quality.
- Experts warn this issue might cause lawmaking based on invented or false evidence.
Read the Full Article
This is a fact-based summary from The Actual News. Click below to read the complete story directly from the original source.
