In August 2025, the Productivity Commission (PC) released its interim report on harnessing data and digital technology.
This is one of five interim reports on productivity (see here). The PC is seeking responses by 15 September.
The interim report covers a range of issues other than copyright, such as privacy and financial reporting. It does not make recommendations about copyright, but does ask for feedback on:
- Are reforms to the copyright regime (including licensing arrangements) required? If so, what are they and why?
- Should Australia introduce a text and data mining (TDM) exception? If so, what are the costs, benefits and risks, and what conditions or limitations should apply?
A TDM exception would cover:
- AI model training, including for Generative AI
- ‘all forms of analytical techniques that use machine-read material to identify patterns, trends and other useful information’, for example to ‘produce large datasets that can be interrogated through statistical analysis’
In seeking feedback, the PC notes its views that:
- large AI models (like Chat GPT) are currently available to be used in Australia : introducing a TDM exception is unlikely to affect that availability
- large AI models are currently trained overseas, not in Australia: it is unclear whether a TDM exception would change this trend
- a TDM exception could affect development of smaller, low compute models in Australia: e.g. by Australian research institutions, medical technology firms, research service providers
Copyright Agency’s position
Australia should not introduce a TDM exception because it is:
- Unnecessary: developers can already get legal access to content in Australia for AI development, including under licensing arrangements
- Harmful: would reduce investment in content, and also bad for developers: who need content for future enhancements