Senior figures overseeing federal parliamentary committees have raised concerns about the increasing prevalence of artificially intelligent material contaminating the policy development process. Investigations conducted earlier this week revealed that inquiries are receiving substantial volumes of AI-created content, including invented academic studies and false attributions to established researchers and scholars. Some committee reports have incorporated citations from submissions where the majority of referenced sources appear to be products of artificial hallucination—instances where language models generate or reference content that appears authentic but has no factual basis.
Liberal Senator Paul Scarr emphasized the significant risks posed by such submissions, noting that decision-makers across government, business, and other sectors face potential harm from basing choices on inaccurate information. He referenced a 2023 incident in which a committee received a submission containing fabricated allegations against KPMG, which the consulting firm disputed and an academic later acknowledged responsibility for. Scarr called for enhanced scrutiny of citations, footnotes, and claims contained in submissions, stating that parliamentary institutions must be vigilant about source verification.
Labor MP Louise Miller-Frost, who chairs the social policy and legal affairs committee, acknowledged the growing awareness of the problem while cautioning against dismissing AI-assisted submissions entirely. She noted that committees and their supporting secretariats have become increasingly conscious of the issue, though the volume of incoming material makes comprehensive verification challenging. Miller-Frost also highlighted potential benefits, suggesting AI tools could assist members of the public who lack resources to compose submissions independently, particularly in social services areas where individuals may have time or capacity constraints.
Senate President Sue Lines indicated that committee members routinely validate information presented during inquiries and that submitters bear responsibility for accuracy. She suggested parliamentary practices would continue adapting to widespread AI adoption, balancing increased public participation against resource constraints and quality assurance demands. Senate guidance on inquiry submissions explicitly cautions that AI use carries risks to information quality and designates submitters as accountable for accuracy. This advisory comes following a separate incident where consultancy firm Deloitte provided a partial refund to the federal government after its AI systems produced fake references and citations in a report valued at $440,000.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI