A report that informed Australia's social media ban for under-16s reportedly contains citation errors that appear to be AI hallucinations. The Age Check Certification Scheme (ACCS), a UK-based organisation that ran the $3.48m age assurance technology trial last year, conceded that ChatGPT was used in editing the report but denied that AI generated the cited materials.
The report was heralded by communications minister Anika Wells as showing "many effective options" for checking ages, paving the way for the ban to come into effect in December last year. However, a submission to an ongoing Senate inquiry highlighted at least two citation errors, suggesting they "appear to be AI hallucinated, rather than being based on real sources."
Guardian Australia's analysis identified six references in one section of the report with errors. These included Digital Object Identifiers (DOIs) linking to non-existent papers or incorrect papers, and author, journal, and publication year combinations that did not match known references. One article cited by ACCS was reportedly accessed in March 2025, but a lead author confirmed it was not publicly accessible until June 2025.
ACCS initially denied using AI in the report but later conceded it was used to rewrite paragraphs more succinctly. They maintained that all links were checked and verified by human verification. Department officials stated in a Senate inquiry hearing that ACCS claimed errors were due to links breaking that had previously worked, though officials have not independently verified this.
The federal government has previously taken a dim view of contractors using AI in reports, with Deloitte refunding part of a contract last year after admitting AI use and errors. Independent senator Fatima Payman called for an apology and refund from the contractor behind the report.