A report backing Australia's proposed teen social media ban contains fabricated academic citations, according to Guardian analysis. The report's authors admit using ChatGPT for editing but deny AI generated the false references.
A comprehensive review of the technical report supporting Australia's social media age restrictions has revealed troubling inaccuracies that cast doubt on its credibility.
The Guardian's investigation identified multiple citations to academic articles that do not exist within a key section of the document. These false references represent a serious concern for a report influencing major policy decisions.
The report's authors acknowledged using ChatGPT during the editing process but rejected claims that AI-generated hallucinations produced the citation errors. Hallucinations—when AI systems confidently generate false information—represent a known weakness of large language models like ChatGPT.
The revelation raises significant questions about the reliability of AI tools in policy research, particularly when government decisions affecting millions of users depend on accurate supporting evidence. Australia has been moving toward implementing restrictions on social media access for users under 16, making the quality of supporting research critical.
The discrepancy highlights the tension between adopting AI efficiency tools and maintaining academic rigor. While ChatGPT can streamline editing workflows, its use in research documentation requires careful verification of all factual claims and citations.
This incident occurs amid growing scrutiny of AI-generated content across institutional settings. Educational institutions, news organizations, and government bodies have grappled with establishing protocols to prevent unreliable AI output from influencing important decisions.
The Senate has requested responses from the report authors regarding the citation issues and their vetting procedures. The controversy may influence how policymakers approach AI tool usage in future legislative research and whether stricter verification protocols become standard practice for government-backed reports.
Alibaba's Qwen 3.8 27B model delivers strong performance across benchmarks, but exhibits a tendency toward excessive reasoning that impacts practical usability.
Promise, a new AI-powered studio based near Sony Pictures in Culver City, represents a shift in filmmaking as creators embrace artificial intelligence to bypass traditional studios and reduce production constraints.
Fields Medalist Timothy Gowers observes that large language models have primarily solved famous mathematics problems by finding counterexamples rather than constructing formal proofs.
A secondary market for AI API credits is developing, with traders buying and selling unused computational allocations. The practice raises questions about pricing efficiency and market dynamics in the AI services sector.