OpenAI has used an appearance before a parliamentary inquiry on AI to apologise again to Australia about the hack on Medicare last month, as media and entertainment organisations argued weakening copyright law for AI model training would leave artists as “roadkill”.
Meanwhile Anthropic told the inquiry it was not seeking to dictate copyright law on model training, but the current rules made it “impossible” to train AI in Australia.
Tuesday morning’s hearing of the joint parliamentary inquiry on artificial intelligence largely focused on copyright licensing for AI training in Australia, amid concerns companies such as Anthropic were pushing the Australian government to adopt an opt-out model, requiring rights holders to specifically opt out of allowing AI to train on their content.
Anthropic’s special envoy, the former US ambassador to Australia Jeffrey Bleich, told the committee that Anthropic had “never tried to dictate to Australia” on copyright.
“We’ve just shared information about how the training works, and then discussed what its implications would be with respect to our capacity to operate here,” he said.
Anthropic suggests current Australian laws mean AI companies need to obtain a licence for the use of every piece of content on the internet as training material. Bleich said this was “technically impossible”.
Anthropic was keen to explore what the Australian government wants to do to incentivise training in Australia, and Bleich said Australia could shape frontier models if they were trained here, otherwise it would ultimately rely on models trained elsewhere.
Sign up for the Breaking News Australia email
OpenAI apologised for its AI agents attacking a number of Australian government websites in June. The multibillion-dollar company notified the government via an email to a public-facing address three months after the breach.
“I want to begin with an apology. During internal training and evaluation, our models accessed Australian government websites in ways they were not directed to. That should not have happened. We also should have handled our response better,” OpenAI’s chief strategy officer, Jason Kwon, told the committee on Tuesday afternoon.
“We are sorry, and we know we have work to do to rebuild trust with the Australian people. We are committed to doing that work.”
Following the revelation of a separate hack on the NSW national parks and wildlife service last week, Kwon told the committee OpenAI was reviewing logs of AI agent training dating back to November 2025, which identified the latest hack activity taking place in June.
He promised that should more incidents be identified, government agencies would be notified “very, very quickly”.
OpenAI representatives, while apologetic, downplayed the agent activity on Australian government websites as not “not super sophisticated” compared with the Hugging Face attack. The independent senator David Pocock questioned why the company didn’t contact ministers, instead sending an email “to some arbitrary department email” after the Medicare incident.
“I think that on retrospect, we should have done what you were suggesting … I think people were thinking about this as a technical situation, and they wanted to contact the technical counterparties – but it’s not good enough,” Kwan said.
Anthropic’s head of safeguards, Dave Orr, told the committee the company had reviewed hundreds of millions of transcripts of its models and had not found any cases in which it interacted with Australian government systems in an unauthorised way.
He could not say whether Anthropic’s customers may have used the models on Australian government systems, given the company’s “zero data retention” policy.
skip past newsletter promotionFree newsletter |When needed
Sign up to Breaking News Australia
Get the most important news as it breaks
after newsletter promotion
Anthropic says AI agents didn’t breach Australian government websites – video‘Artists will become roadkill’
Annabelle Herd, the chief executive of the Australian Recording Industry Association (Aria), told the committee that “Australia’s artists will be the roadkill in the rush” to make deals with AI companies, and said Aria was not against AI frontier model training being done in Australia but it should be done on “our terms”.
“There is nothing less sovereign than rewriting our laws to suit foreign company commercial interests, done the way that big AI is proposing,” she said. “It will devastate the music industry and make Australia a pariah for rights holders globally.”
She said Anthropic and OpenAI were “asking the government for a free ride” and the creative sector was calling on the government not to cave in to their demands.
The ABC told the committee Australia’s existing copyright regime was “completely adequate”.
The ABC’s head of content and legal operations, Kate Gilchrist, said an opt-out system as proposed by AI companies would place burdens and costs on rights holders, as it would require ongoing surveillance to see whether copyright material was being used.
She said for the ABC, it would mean ensuring every site on which the ABC publishes material had opted out.
“We cannot scour the internet and ensure we are opting out on all of those different sites of publication. It simply does not work,” she said.
Gilchrist said she suspected ABC content had been scraped already.
The union representing the entertainment and news sector, the Media Entertainment and Arts Alliance (MEAA), has called for an equitable remuneration scheme that would put fees on companies across the AI technology stack.
“Revenue would support two things: direct payments to creative and media workers whose work contributes to AI systems. And secondly, a cultural fund supporting Australian live entertainment, screen music, public interest journalism and cultural activity,” the MEAA’s acting chief executive, Adam Portelli, said.
The committee is holding hearings from Tuesday to Friday this week.