The heads of a parliamentary investigation into AI adoption in Australia have denied there was government interference into early recommendations they issued in a letter to the Prime Minister overnight.
Labor chair Jo Briskey and Liberal deputy chair Jonno Duniam have defended the letter sent from their bipartisan parliamentary committee after Independent Senator David Pocock raised concerns the prime minister’s office had helped draft it.
It urged the government to act immediately on four key areas, including strengthening AI safeguards, building sovereign AI capability, fixing copyright settings, and having nationally consistent data centre standards.
Senator Duniam said the committee had issued advice ahead of the final report which is due by November 30 because they understood AI development was moving rapidly and the government had limited time to act if Australia wanted to pursue AI sovereignty.
A circulated draft copy of the letter, however, had revealed notes and marking to amend before it was published.
Senator Pocock said it raised concerns that the committee already has a predetermined outcome and was just a vehicle to serve a political purpose rather than parliamentary scrutiny and policy input.
Greens leader David Shoebridge echoed Senator Pocock’s concerns and labelled the committee a “farce”.
“The farce that that committee has become so quickly, with their letter to the prime minister having been vetted and edited,” he said.
“This is not some kind of independent parliamentary committee.
“This is just some kind of highly expensive choreographed rubber stamp for the prime minister.”
Senator Shoebridge labelled the recommendations in the letter, which was sent on Monday evening, as things which could have been “pulled out of the bottom drawer three years ago”.
The recommendations included embedding independent oversight inside AI companies, give security agencies greater access and improve literacy, skills, and adoption.
But during a press conference and TV interview on Tuesday, Senator Duniam hit back at accusations it had been a “stitch up” — insisting there was “no guidance, no interference”.
The Liberal Senator admitted he penned the letter and that he welcomed the editors after “typographical and grammatical errors” were picked up.
“I wrote the letter. It was replete with typographical and grammatical errors, so I’m pleased others picked it up,” Mr Duniam told News24.
He and Ms Briskey, who were signatories on the letter, told Canberra reporters that they had held a late-night meeting with fellow members to finalise the letter which they had reached a consensus about.
Ms Briskey was asked the criticisms over the involvement by the Prime Minister’s office and denied the early advice translated to predetermined recommendations in the final report.
“It’s been a process by which we’re seeking advice and making sure that this is working across the board. There was engagement right across the parliament,” she admitted.
“Senator Duniam also worked with his side, the opposition.
“This is about the parliament doing the work on getting an outcome and progressing this for Australians.
“We understood that there was consensus across some key things around safety, transparency, and accountability that we can put to government to pursue at pace and to keep ourselves ahead of this.
“There is more work to do for this committee. We have not finished our work.
“This was just an early indication of where we think things are heading.
“There is a full report that will go to the Parliament before the end of this year, giving everyone across the committee the opportunity to contribute to final recommendations.”
The committee heard from tech giants OpenAI and Anthropic in hearings last week after Sam Altman’s company admitted that its autonomous AI agents gained unauthorised access to Medicare statistics via Services Australia in June.
Ms Briskey said the committee wouldn’t be shy to recall the Silicon Valley tech giants if needed before the end of the year.
“It is absolutely open to the committee to consider doing, and if that’s the case, then that may be what we seek to do,” she said.
Both Ms Briskey and Senator Duniam warned Australia needed to act swiftly to “get the settings” right to “keep Australians safe” and “play in the global response to artificial intelligence”.
“That is an opportunity that’s open to Australia right now, but it won’t be open forever,” Ms Briskey warned.
“We’ve got to make sure we walk through this window of opportunity but do so on our terms.”
Senator Duniam described it as the “biggest issue confronting our nation” and said a “dynamic response” was important.
When asked around the timeline, Senator Duniam said he felt the early advice issued in the letter needed to be acted on “this year”.
“This is the biggest issue confronting our nation at the moment. I don’t think a lot of people have a handle on that,” Senator Duniam said.
“There’s been no guidance, no interference. There has been a lot of discussion, a lot of consultation.
“We have to get it right. We have to do it on Australia’s terms. Those who want to be constructive can. Those who want to sit outside and cut from the sidelines and find points of difference will, but we’ve got to get this right.”
Mr Albanese told Parliament on Tuesday that the AI wave was a “revolution” which would have “an enormous impact on the way society functions”.
He called for the parliament to work together and to “all engage as much as possible”.
“It is a revolution that will have an enormous impact on the way society functions,” he said.
“I’d urge the entire Parliament to engage on these issues.
“It certainly should be and not be an area of partisanship.
“It needs to be something in which we all engage as much as possible.”
The PM later addressed concerns around AI developers using Australian content to train their models during an appearance on ABC’s Afternoon Briefing.
He warned Australia’s not to be naive to believe that the scarpapping of intelligence data was already occurring.
“We can’t just pretend that it isn’t happening,” he told Afternoon Briefing.
“This is a complex issue. We need to get on top of it.
“AI has an enormous potential to boost productivity, to boost business efficiency, to create jobs.
But Mr Albanese added that the Medicare statistics breach was a “lesson” for Australia of the risks but also for why the country should help shape how AI is adopted.
“We need to make sure that we act in the interests of humanity so that we are in control,” he said.
“We need proper regulation of it. Engagement with the frontier companies is important. That needs to be something that nations can play a role in.”
The letter noted that existing copyright rules are failing to secure licensing deals for Australian content and that local creators should have proper control and compensation.
However it also acknowledged that major frontier AI developers viewed Australia’s current copyright framework as a roadblock to training locally.
“Frontier AI labs indicate existing copyright settings are a barrier to undertaking training in Australia,” it stated.
The letter also urges the government to build a competitive Australian AI industry spanning data sovereignty, critical infrastructure, skills, procurement, industry development, public-interest compute, and Pacific partnerships.
The Productivity Commission and Australian Electoral Commission are listed to appear before the committee on Wednesday.
Get the latest news from thewest.com.au in your inbox.
Sign up for our emails