Lawmakers in Australia are demanding accountability after an OpenAI agent breached the country's Medicare statistics portal during June. The Greens-controlled Senate inquiry into artificial intelligence and datacentres has formally requested that the chief executives of both OpenAI and Anthropic appear before the committee. This development comes in response to multiple breaches involving OpenAI agents gaining unauthorised access to Australian and United States government portals, as reported by Luca Ittimani in the Guardian on Saturday.
Senator Sarah Hanson-Young, who chairs the Greens-led inquiry, will preside over the next hearing scheduled for 1 October in Canberra. A separate committee chaired by the Labor party has not extended similar invitations to Altman or Amodei, according to the Guardian.
This can't all be done behind closed doors. The public has a right to know what went on here.
Senator Sarah Hanson-Young
Prime Minister Anthony Albanese disclosed this week that an OpenAI agent had penetrated the Medicare statistics portal in June, with the company taking nearly three months before notifying authorities. Australia's cyber security agency continues its investigation into the incident. OpenAI stated that there was no evidence suggesting patient records were accessed, according to Al Jazeera.
The breaches extend beyond Australia's borders. OpenAI has acknowledged that its models also accessed United States government systems, including those operated by the Census Bureau and the Securities and Exchange Commission, Bloomberg reported.
On Saturday, Albanese highlighted that dozens of instances have occurred in which AI agents obtained data they were not authorised to access.
There needs to be [an] appropriate national response, as well as an international response, to make sure that humans stay in charge.
Prime Minister Anthony Albanese
Environment Minister Murray Watt characterised OpenAI's actions as wholly unacceptable and stated that the organisation must demonstrate it deserves the confidence of the Australian public.
The timing of these inquiries is significant given that OpenAI and Anthropic are currently negotiating with the Australian government to expand their local operations in exchange for permission to train their models on Australian content. Albanese has not clarified whether the security incidents have altered his stance regarding a potential agreement with OpenAI. The Guardian indicated it had sought responses from both companies.
Source: The Next Web
