Anthropic, OpenAI call for Australia to relax ban on training of AI models

SYDNEY, Sept 22 : Top AI companies OpenAI and Anthropic urged Australia to rethink a ban on letting them use the country's creative content for training their models, saying a limited exemption could protect artists' rights while encouraging more investment.

Countries around the world are scrambling to settle on how to regulate the explosion of AI while courting investment that economists say may be worth hundreds of billions of dollars by the end of the decade.

Australia has said it will implement a broad set of rules on AI and data centres to take effect from next year, but has already ruled out an exemption to copyright laws which effectively prevents the mostly US-domiciled companies from undertaking training there.

In submissions to a parliamentary inquiry in September on laws governing AI, the companies acknowledged the ban, but called for it to be relaxed.

"Anthropic accepts that broad exceptions from copyright ... have not received widespread support and have been ruled out by the government," said the owner of AI chatbot Claude in its submission.

"As an alternative, we believe that the government could permit AI model training through a narrow form of conditional approval," it added.

If Australia required "investment or other conditions designed to support the future of Australian creators and cultural endeavours ... we are open to considering those conditions," it said.

ChatGPT maker OpenAI called for "a balanced copyright framework that allows models to learn from publicly available information" while offering rightsholders opportunities to collaborate.

Both companies tied the issue to major infrastructure investments underway in Australia, with OpenAI having signed an offtake agreement with Australian data centre developer NextDC for a planned data centre in Sydney and Anthropic named last week as a partner for a data centre in Queensland.

The submissions add to pressure on the government as it weighs the interests of AI developers against those of authors, publishers and other creators, who have resisted exemptions allowing their work to be used to train commercial models without consent or payment.

The Joint Select Committee on Artificial Intelligence is due to deliver its report in November.