(New York, NY) — Australia’s Prime Minister says an artificial intelligence agent belonging to OpenAI went rogue and hacked into an Australian government health website, gaining access to non-public information. Anthony Albanese [[ Al-buh-nee-zee ]] says the breach occurred in June but was only disclosed last month. Both Albanese and OpenAI CEO Sam Altman were […]
Science
PM Says OpenAI Agent Hacked Australia Government Website
Audio By Carbonatix
(New York, NY) — Australia’s Prime Minister says an artificial intelligence agent belonging to OpenAI went rogue and hacked into an Australian government health website, gaining access to non-public information.
Anthony Albanese [[ Al-buh-nee-zee ]] says the breach occurred in June but was only disclosed last month.
Both Albanese and OpenAI CEO Sam Altman were in New York for the United Nations General Assembly.
Albanese says he also expressed his disappointment that notification took “way too long,” and said the nature of the notification was “unacceptable.”
Albanese says the breach occurred June 18th, but that OpenAI only notified the country on September 10th — in an email sent to a public mailbox.
OpenAI reportedly discovered the unauthorized activity in August. It’s believed that only non-sensitive information was accessed.
Albanese says he spoke with OpenAI’s CEO to express his “extreme concern” about the incident.