Back to Articles
AI Companies Must Take Swift Action: Andrew Charlton

ABC News

READ

Details

Date Published
7 Oct 2024
Priority Score
4
Australian
Yes
Created
7 Oct 2026, 04:00 am

Description

The company behind the hacking of Australia's Medicare website - Open AI - says it's changed its systems to ward off future breaches by rogue agents.

Summary

The Australian Government is demanding increased transparency and faster incident reporting from OpenAI following a security breach involving the hacking of Medicare data. Assistant Minister Andrew Charlton emphasizes that the upcoming Australian AI standards will mandate safety, transparency, and disclosure requirements to mitigate risks to the public. The establishment of the Australian AI Safety Institute is highlighted as a critical technical capability for testing and analyzing frontier models for safety before they cause significant harm. This development marks a significant move toward legislative governance of AI capabilities to ensure they do not outpace public trust or safety frameworks.

Body

The company behind the hacking of Australia's Medicare website - Open AI - says it's changed its systems to ward off future breaches by rogue agents. The Assistant Minister for Science, Technology and the Digital Economy says the company must ensure it swiftly reports any breach.More InformationFeatured: Andrew Charlton, Assistant Minister for Science, Technology and the Digital Economy CreditsRachel Mealey, AuthorImage DetailsAssistant Minister for Science, Technology and the Digital Economy Andrew Charlton (ABC News: Matt Roberts )Program:More from AMTranscriptRachel Mealey: The company behind the hacking of Australia's Medicare website, OpenAI, says it's changed its systems to ward off future breaches by rogue agents. Earlier I was joined by the Assistant Minister for Science, Technology and the Digital Economy. Andrew Charlton, thanks for speaking to AM.Andrew Charlton: Good to be with you.Rachel Mealey: Was OpenAI's apology to Australia full-throated enough?Andrew Charlton: Well, what we learned yesterday was that Australia was right to call out the hacking that occurred. And we also learned that the action the PM took is already having an impact on how these companies approach safety. We take OpenAI's apology at face value, but we don't rely on it. Australia is taking its own action to make sure that we keep Australians safe from the risks of AI. And those actions will include the new Australian AI standards, which will go directly to issues of safety, transparency and incident disclosure, which the PM has said will be out before the end of the year.Rachel Mealey: OpenAI's Chief Strategy Officer, Jason Kwon, has promised that should more incidents be identified that government agencies would be noted, quote, very, very quickly. Are you confident that the reporting lines are now up to scratch?Andrew Charlton: Well, clearly, the reporting that OpenAI did in relation to the last incident was inadequate. It was inadequate in its timeliness and it was inadequate in its nature. The Prime Minister has launched a rapid review of that incident. And one of the outcomes of that review will be to inform the government's new AI standards, which will govern exactly how this type of incident will be reported. And clearly, we need to make sure that those standards require these companies to report in a timely way and in an appropriate way. And both of those things will be covered in the new AI standard.Rachel Mealey: Do you accept his position that his boss, OpenAI's boss, Sam Altman, didn't know about the breach when he met in person with Australia's Deputy Prime Minister, Richard Marles, on September the 1st?Andrew Charlton: Well, I don't have any information otherwise. But what I can say is that it is inappropriate for a company like this to have such a long period of time between when they conduct a hack like this and when they notify the Australian Government. I'm not privy to their internal communications, but what we have said to them very clearly is when incidents occur, they need to front up in a timely way and provide appropriate notification so the Australian Government can address these issues.Rachel Mealey: Jason Kwon basically said that his company would do its part to keep legislators up to date with the technology, but that governments also needed to step up to meet that technological know-how. Is there that depth of knowledge in the Australian Government?Andrew Charlton: Well, the Australian Government has acted quickly to address the risks of AI safety. We established the Australian AI Safety Institute. We are one of a small number of countries in the world that has one of these capabilities within its government. That capability is exactly what you describe. It's a group of technical experts who are able to test and analyse these models for their safety to make sure that where there are risks to Australians, we are mitigating those risks. We've also announced our government's safety priorities, which cover privacy, transparency, automated decision making. And now the government is moving forward with its Australian AI standards, which will legislate many of these principles. CanRachel Mealey: they all keep up with the pace of change, though?Andrew Charlton: Well, things are moving very quickly in the world of AI, but we need to make sure that they don't move more quickly than public trust and public safety. And that's why the government has been working to ensure that we introduce these new standards. These are world first standards to ensure safety, transparency and disclosure around incidents like the one we saw on Medicare.Rachel Mealey: Does it all highlight to you that Australia can't do anything without the help of the US? And at the moment, the Trump administration isn't listening.Andrew Charlton: No, Australia has significant agency here. I mean, I would point out that the Prime Minister called out this behaviour. He was the first leader in the world to call out this hacking behaviour. And that has had consequences and repercussions that has caused these companies to change their behaviour. We'll also be the first country in the world to introduce the Australian AI standards, which these companies have, many of them have signed up to the principles underneath. So we have the power to influence the way that artificial intelligence comes into Australia. And we're determined to do that in a way that makes sure that it benefits Australians and doesn't cause them harms.Rachel Mealey: Andrew Charlton, thanks for joining AM.Andrew Charlton: Thanks very much.Rachel Mealey: And Andrew Charlton is the Assistant Minister for Science, Technology and the Digital Economy. Appears In Jewish mark October 7 anniversaryDuration: 20 minutes 18 seconds20m 6itemsIn this episode1 of 6Israeli's mark October 7 anniversaryDuration: 4 minutes 17 seconds4m 17s2 of 6Vigil for October 7 held at Bondi BeachDuration: 5 minutes 33 seconds5m 3 of 6PlayingAI companies must take swift action: Andrew CharltonDuration: 5 minutes 4 seconds5m 4 of 6University groups slam Coalition migration planDuration: 3 minutes 26 seconds3m 26s5 of 6French student protests turn violentDuration: 2 minutes 25 seconds2m 25s6 of 6Ozempic-like drugs pose 'serious' surgery riskDuration: 3 minutes 16 seconds3m 16s