Context

On 26 March, 2024, the Senate called for the establishment of a report by the Select Committee on Adopting Artificial Intelligence. The committee accepted written submissions by 10 May 2024, and was concluded on 26 November, 2024.

Final Report and Recommendations

In the final report, unlicensed training of large language model’s (LLM) was described as an “unprecedented theft”, underscoring the need for prior consultation, transparency, a pathway to licensing, and mechanisms for remuneration. The final report formulated 13 recommendations, which are summarised below:

  1. Legislation for high-risk AI regulation
  2. Principles-based approach to defining high-risk AI
  3. Inclusion of general-purpose AI models in high-risk list*
  4. Support for sovereign AI capability in Australia*
  5. AI risks to workers’ rights
  6. Extend work health and safety laws
  7. Consultation on AI’s impact on workplaces
  8. Consultation on AI’s impact on creative industries
  9. Transparency and licensing for copyrighted works in AI training*
  10. Fair remuneration for AI-generated outputs using copyrighted material*
  11. Implement privacy act recommendations on automated decision-making
  12. Implement recommendations for a legal framework and monitoring body
  13. Holistic approach to ensure sustainability of  AI infrastructure in the national interest of Australians

*Highlighted by APA.

These recommendations were welcomed by the Australian Publishers Association (APA) as these recommendations align with APA’s policy on Artificial Intelligence.

Originally published by APA here.

Find the final report here.