Beta The Briev beta is out. Free on iPhone via TestFlight — install it in under a minute.

Join the beta ↗
Briev
Live
Politics

AI-Generated Errors Mar Government Report on Australia’s Under-16 Social Media Ban

A report on age-verification technology for Australia’s teen social-media ban was found to contain AI-created citation errors, prompting Senate scrutiny.

The Age Check Certification Scheme conducted a $3.48 million pilot of age-assurance technologies intended to support Australia’s ban on users under 16, with communications minister Anika Wells citing the study as evidence of viable options. However, a Senate inquiry uncovered citation problems, noting that several references in the report seem to be AI hallucinations rather than genuine sources. Guardian Australia’s review found six DOI links that either pointed to non-existent papers, the wrong articles, or mismatched bibliographic details, including a study accessed before its publication date and a misattributed author named “Jamil.”

ACCS later conceded that ChatGPT was used to edit paragraphs but insisted the references were manually verified, though their own corrective list also contained errors. Department officials said the links worked when the report was released and that any broken links likely changed later, while experts warned that such inaccuracies could undermine policy decisions. Senators, including Fatima Payman, called for accountability and possible refunds from contractors.

Why it matters

Faulty AI-generated citations could lead to poor policy choices on youth online safety.

How the sides frame it

LOW AGREEMENT

Left-leaning coverage highlights AI-generated citation errors that call the government report’s credibility into question, while center coverage criticises the teen social-media ban as ineffective and politically rushed.

LEFT

Exposes AI-hallucinated references that undermine the report’s reliability

CENTER

Attacks the ban’s lack of results and the government’s hurried promotion of the policy

The left emphasises

  • several references appear to be AI hallucinations
  • ChatGPT was used to edit paragraphs
  • six DOI links point to non-existent or mismatched papers

In this story

teen social media banAI hallucinationsChatGPTcitation errorsSenate inquiry
Get the beta ↗