An OpenAI model hacked an Australian government website earlier this year, marking one of the first known cyberattacks by AI on a government database.
The company took three months to notify Canberra that its AI system gained unauthorized access to files on a website for reporting healthcare statistics, Prime Minister Anthony Albanese said Thursday Australian time. The PM, speaking from New York, said he had spoken with OpenAI Chief Executive Officer Sam Altman to express "extreme concern about this incident" and the length of time before notification.
Also Read | Layoffs Are Rising In Tech, But These Skills Are Helping People Unlock Better Pay
There's no evidence that the personal information of Australian citizens was compromised, Albanese said, adding that the government is now investigating whether any crimes were committed. OpenAI said that between discovering the breach in August and notifying the Australian government on Sept. 10, it was validating and investigating the facts and what information had been accessed.
The hack occurred on June 18, adding to a series of incidents this summer where OpenAI models broke into online resources of outside companies, most notably with the case of Hugging Face, a website hosting AI datasets. OpenAI sent its notification to the Services Australia public inbox in early September, Albanese said.
ALSO READ: Temporary Chats To Model Training: What's Inside ChatGPT's New Privacy Center?
"I also expressed my disappointment that it took the company way too long to inform the government what had occurred," the Australian leader said at a press conference on the sidelines of the United Nations General Assembly. "The nature of the way that the notification occurred as well was unacceptable."
This is the first known time that an artificial intelligence model has gained access to Australian government systems, Defense Minister Richard Marles said on radio. The model accessed four state and federal government websites requesting data, he said, and when the Services Australia website declined to provide the information, it "effectively hacked into that medical portal and got that information anyway."
"We will look at what is the legal situation in respect of this and what it means to have gained an unauthorized access, albeit in an unintended way," Marles, who is also deputy prime minister, said.
A forensic examination, aided by the Australian Signals Directorate, is underway to ascertain more information about the breach.
In a statement, OpenAI said its "extensive review of misaligned model activity" had identified actions involving "several Australian government websites and services," which its models had accessed when they attempted to look up statistics to answer questions about Australia during an internal evaluation.
"In the course of that, our models took actions we did not intend," a spokesperson for OpenAI said in the statement.
The AI system was assigned to find information and "there were blocks clearly which were coming back telling the AI agent, no. The AI agent found a way around those blocks," Albanese said. "The model attempted alternative ways to obtain the info that it wanted, and this led to unauthorized access into some other areas."
Albanese said he needed to be reassured that the correct protocols to manage AI were in place, adding that the government's "review found no evidence of patient records being accessed. The information accessed included aggregate health statistics and internal file names."
"Mr. Altman certainly has acknowledged - and I've invited further discussions with him as well - that their protocols were not up to scratch here," Albanese said.
ALSO READ: 'No Real Moats': The Big Short's Steve Eisman Says AI Giants Are Faking A Crisis To Stop Competition
The Australian leader's comments come as OpenAI urges the US to lead a global effort to set standards for AI. Earlier Wednesday in New York, Altman and Anthropic PBC CEO Dario Amodei called on world leaders to work together in a presentation to the United Nations Security Council that focused on rising concerns the technology poses existential risks to humanity.
Anthropic's technology has also figured in undermining online security, with a hacker last year using its Claude model to carry out a series of attacks against Mexican government agencies, resulting in the theft of a trove of sensitive tax and voter information, according to cybersecurity researchers. Amodei and Altman have both expressed concern about the potential threat of self-improving AI systems outrunning human oversight and control.
(This story has not been edited by NDTV staff and is auto-generated from a syndicated feed.)
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.
