Labor blocks AI industry moves to waive copyright

The federal government won’t change the law to allow tech companies to train artificial intelligence (AI) models using copyright material, Australian media outlets reported today.

The decision is a rejection of the Productivity Commission’s interim recommendation earlier this year that copyright rules be relaxed to enable data harvesting by tech companies. The rejection is already being welcomed by media, publishing and bookselling bodies.

“Australian creatives are not only world-class, but they are also the lifeblood of Australian culture, and we must ensure the right legal protections are in place,” attorney-general Michelle Rowland said on Sunday.

As Books+Publishing reported in August, the Productivity Commission’s interim report was met with strong opposition from author organisations and the publishing and recording industry. A coalition of industry organisations and authors addressed the Senate earlier this month, with author Tom Keneally stating, ““It’s not copy-charity. It’s not copy-privilege. It’s not copy-indulgence. It’s copyright. And our right has been taken away by ignorant people who don’t realise what copyright is.”

The attorney general will convene the government’s copyright and AI reference group on Monday and Tuesday to examine other options for responding to AI’s use of copyright material. A key challenge is to identify when tech companies have in fact used copyright material to train AI.

“Under the reforms being considered by Rowland,” wrote the Australian Financial Review’s Ronald Mizen, “this would be resolved by forcing tech companies to disclose the content used to train their models. This could also be supplemented by a new enforcement regime that makes it easier for small rights holders to seek compensation for their work.”

One option is for a collective process similar to the licensing function of the Copyright Agency.

Photo: Attorney-general Michelle Rowland

Category: Local news