{Economic Daily} reported exclusively on the 24th that Australian Prime Minister Albanese stated that an AI agent from OpenAI breached Australian government websites, but the company only notified the Australian government nearly three months after the incident occurred.
The incident led to a breach of Australia's Medicare website. Albanese said he spoke with OpenAI CEO Altman in New York on Wednesday (23rd), expressing "extreme concern" about the incident.
The breach occurred on June 18. OpenAI discovered the issue in August, and the Australian government was not informed until September 10, when OpenAI sent an email to the public mailbox of Australia's Department of Civil Services.
Albanese, speaking at a press conference in New York during the United Nations General Assembly, said: "It took OpenAI far too long to inform the government about what had happened. I am disappointed by this, and such a notification method is unacceptable."
In a statement, OpenAI said the company conducted a broad review of model behaviors that did not meet expectations and found its models involved "multiple websites and services of the Australian government." Between the time OpenAI discovered the breach in August and notified the Australian government on September 10, the company had been verifying and investigating the facts and confirming what information was involved. No access to patient records was found, and the information involved included health statistics and internal documents. (rc)
The incident led to a breach of Australia's Medicare website. Albanese said he spoke with OpenAI CEO Altman in New York on Wednesday (23rd), expressing "extreme concern" about the incident.
The breach occurred on June 18. OpenAI discovered the issue in August, and the Australian government was not informed until September 10, when OpenAI sent an email to the public mailbox of Australia's Department of Civil Services.
Albanese, speaking at a press conference in New York during the United Nations General Assembly, said: "It took OpenAI far too long to inform the government about what had happened. I am disappointed by this, and such a notification method is unacceptable."
In a statement, OpenAI said the company conducted a broad review of model behaviors that did not meet expectations and found its models involved "multiple websites and services of the Australian government." Between the time OpenAI discovered the breach in August and notified the Australian government on September 10, the company had been verifying and investigating the facts and confirming what information was involved. No access to patient records was found, and the information involved included health statistics and internal documents. (rc)