Senators are demanding that the chief executives of OpenAI and Anthropic face a Senate inquiry after rogue OpenAI agents hacked Australian and US government websites.

On a visit to New York on Thursday, Prime Minister Anthony Albanese revealed details of an unauthorised OpenAI infiltration of the Medicare Statistics Reporting portal.

OpenAI later confirmed its bots had infiltrated US government websites and leaked more than 50 ChatGPT user photos.

OpenAI founder Sam Altman and Anthropic co-founder Dario Amodei have been requested to appear at the Greens-led inquiry into AI and data centres.

OpenAI CEO Sam Altman attends a state dinner hosted for China’s President Xi Jinping. Photo: AAP

“This can’t all be done behind closed doors – the public has a right to know what went on here,” committee chair and Greens senator Sarah Hanson-Young told the Sydney Morning Herald.

“If they truly believe their own warnings, they must front up, face the Senate’s questions and have an honest conversation about what effective, lasting regulation of this industry should look like,” she said.

A separate joint committee, led by Labor members, has not made a similar request.

On his return from New York on Saturday, Albanese said there had been been dozens of cases of artificial intelligence agents wrongfully accessing classified information beyond Australia’s Medicare hack.

AI was front and centre in Anthony Albanese’s address to the United Nations in New York. Photo: AAP

Yet he did not know about them beforehand.

“At the time we made our announcement … we weren’t aware of other occurrences,” the prime minister told reporters on the tarmac at Sydney airport.

“We were acting purely in our national interest.”

Albanese addressed the United Nations General Assembly in New York on the international need to rein in AI.

He says news that rogue AI agents hacked into US government websites is further proof of that need.

Autonomous agents are known to have infiltrated American university and commerce department sites as well as the US Securities and Exchange Commission.

“We now know it wasn’t just the issue of the Australian sites that have been subject to AI agents accessing information without authorisation,” Albanese said.

“What this does is confirm that approach of making sure that it needs to be an appropriate national response, as well as an international response to make sure that humans are in charge of this new and emerging technology.”

Albanese said it was up to OpenAI to state why multiple breaches occurred.

Jane Hume says the government failed to detect the Medicare website breach. Photo: AAP

Asked to comment on the timing of the Medicare hack reveal and Albanese’s advocacy at the UN, Opposition deputy leader Jane Hume was reticent.

“When it comes to conspiracy theories versus cock-up, I’d always choose cock-up,” she told 2GB Sydney on Saturday.

“The only reason the government detected this breach to our systems was because OpenAI told them that they’d done it.”

The Medicare hack occurred in June but it took until early September for OpenAI to inform the federal government via a public email inbox.

Albanese spoke with OpenAI chief executive Sam Altman by phone while in New York, with the tech boss admitting the company had not done enough.

Australia is one of more than 20 countries to sign a joint statement calling for countries to work together to enact global AI standards and ensure human control.

However, US President Donald Trump used a speech at the UN to lash out at calls to put the brakes on AI.

-with AAP