OpenAI and Anthropic have urged Australia to reconsider restrictions preventing AI companies from using copyrighted creative works to train models, proposing limited exemptions tied to support for local artists and investment.
Australia plans to introduce broader AI and data centre rules next year but has ruled out a general copyright exemption, effectively restricting companies from training models on protected material in the country.
Anthropic, the developer of Claude, proposed a narrow approval system that could allow AI training under conditions requiring investment in Australian creators and cultural projects.
OpenAI called for a copyright framework that allows AI models to learn from publicly available information while giving copyright holders opportunities to work with AI companies.
Both firms linked the issue to their growing infrastructure commitments in Australia. OpenAI has an agreement with data centre operator NextDC for a planned Sydney facility, while Anthropic was recently announced as a partner for a data centre project in Queensland.
The proposals come as Australia weighs competing interests between AI developers and authors, publishers and other creators who oppose allowing commercial AI systems to use their work without consent or payment.
A parliamentary committee examining Australia’s AI laws is expected to release its report in November.
Leave a comment