In what experts claim is the first instance of its sort in history, a rogue OpenAI bot gained access to private information on an Australian government website in June. Prime Minister Anthony Albanese stated in New York on Wednesday, local time, that the agent “infiltrated” a statistics portal that held “non-sensitive” data from Medicare, Australia’s universal healthcare program. He told OpenAI CEO Sam Altman that there would be “legal consequences” for taking “too long” to reveal the breach during a “very frank discussion” with Altman.
OpenAI claimed to have discovered the breach only in August when examining “misaligned model activity” and sent an email on September 10 to an Australian government agency’s general inbox. The email was forwarded to Australia’s cybersecurity center five days later by that government agency, Services Australia, before the prime minister and a government minister were informed.
Albanese claimed that when he spoke with Altman, he expressed “Australia’s extreme concern about this incident” and expressed his “disappointment” that the business had waited months to disclose the breach and “the nature of the way” it did so. According to the Australian leader, Altman had admitted that OpenAI had “issues with protocols.” Albanese stated that the nation’s cybersecurity agency would conduct a “forensic investigation” to see whether additional government systems were impacted.
Also Read:
India’s Corporate Crown Prepares for Change Amid Boardroom Mutiny
In a Security Test, Google’s Gemini AI Compromised Three Businesses
