technology policy•Mon, August 17, 2026•5 min read

Australia’s Teen Social Media Ban Report Faces Scrutiny Over AI Hallucinations

A pivotal report used to support Australia's landmark social media ban for under-16s is currently under Senate review following allegations of AI-generated misinformation. Critics warn that the inclusion of non-existent academic citations risks undermining public trust in critical digital policy.

A Policy Under Pressure

The Australian government’s ambitious attempt to restrict social media access for those under 16 is facing fresh scrutiny. A Senate inquiry has heard that a key report, which provided foundational support for the legislation, contains citations for academic articles that do not appear to exist. The revelation has sparked concerns regarding the integrity of evidence used to shape modern digital regulation.

The report in question stems from a $3.48 million age assurance technology trial conducted by the UK-based Age Check Certification Scheme (ACCS). While the authors of the report have admitted to using ChatGPT to assist in the editing process, they have denied that the erroneous citations were caused by 'AI hallucinations.' However, an independent analysis identified six specific references within the document that contain clear errors.

Impact and Public Trust

Professor Christian Downie of the Australian National University warned that the presence of false citations in government-backed research could significantly erode public confidence. As the government continues to rely on complex technical trials to justify strict online safety measures, the accuracy of that data has become a focal point for critics and legal experts.

The report was tied to a $3.48 million trial on age assurance technology.
Authors admitted to using AI tools for rewriting and editing content.
Multiple academic sources referenced in the document could not be verified.
Social media platforms have expressed ongoing caution regarding the accuracy of early data sets.

If reports or submissions to government are found to have false citations in them – AI or otherwise – it could erode public confidence and trust in institutions.
— Prof. Christian Downie, Australian National University

The Reality of the Ban

Beyond the controversy surrounding the report, the impact of the under-16 ban itself remains a mixed bag. Data from the first few months post-implementation suggests a decline in official account ownership among teenagers, yet many still retain access. Interestingly, there has been a observed shift in behavior, with more children turning toward messaging services and video games. Crucially, the reported frequency of teenagers using AI chatbots has shown no statistically significant change, despite concerns from the eSafety commissioner about these tools helping kids circumvent the rules.

Key Takeaways

A foundational report for Australia's under-16 social media ban is being investigated for containing non-existent academic citations.
The report’s authors acknowledged using ChatGPT for editing but denied that the errors were caused by AI hallucinations.
Experts warn that technical errors in government research threaten the legitimacy of digital safety legislation.
Early impact data suggests that while social media usage among teens has declined, many have simply migrated to gaming and messaging platforms.
There has been no significant change in the use of AI chatbots by teens following the ban's introduction.

FAQ

What is the Australian social media ban?

Implemented on December 10, 2025, the law mandates that social media platforms take reasonable steps to prevent children under 16 from holding accounts.

Why is the recent report being questioned?

A Senate inquiry revealed that the report contains citations for academic papers that do not exist, raising concerns about the accuracy of the data supporting the ban.

Did the authors use AI in their report?

Yes, the authors confirmed that ChatGPT was used for editing and rewriting paragraphs, though they contest that it was the source of the citation errors.

Has the ban been effective?

Early data indicates a decline in social media account ownership, but many children have migrated to messaging apps and gaming instead, with limited impact on AI chatbot usage.