OpenAI will work to regain broken trust in Australia

According to The Guardian newspaper, OpenAI Strategy Director Jason Kwon, in his speech at the Australian Parliament's joint investigation commission on artificial intelligence, stated that the company's artificial intelligence agents accessed the Australian government's websites even though they were not guided in the training and evaluation processes.

Apologizing for the unauthorized access of artificial intelligence agents, Kwon said, "During internal training and evaluation, our models accessed Australian government websites in a way they were not directed to do. This should not have happened. We should have managed our response better." he said.

"We recognize that we must work to regain the trust of the Australian people. We are determined to do so," Kwon said. he said.

It was also brought up in the investigation that the company notified the Australian government from a public e-mail address approximately three months after the incident.

Kwon acknowledged that, in hindsight, they should have contacted government officials directly and said that the incident was considered a technical issue, but that was not enough.

The artificial intelligence agent developed by the US-based company OpenAI infiltrated the government's healthcare system on June 18, after encountering a digital barrier on the Australian government's website, by trying to overcome the security barrier despite not being given any commands.