Regarding the incident where its models infiltrated an Australian government website, OpenAI announced on September 29 its follow-up actions: on one hand, it will provide funding to help Australia strengthen its network defenses, and on the other hand, it will establish a special task force to study the risks brought by increasingly capable AI agents.
OpenAI stated that it will use the "Daybreak for Frontline Defenders" fund, which has a scale of 1 billion US dollars (approximately 6.722 billion Chinese yuan at the current exchange rate), and dispatch a technical team to support Australian local governments and industries in reinforcing the protection of critical infrastructure.
New Task Force, Chief Strategy Officer to Be Questioned in Australia Next Week
Along with this funding, OpenAI will also set up a special working group tasked with drafting actionable policy recommendations regarding "AI agents with continuously improving capabilities," focusing on improving incident reporting mechanisms and enhancing coordination with Australian authorities.
The immediate trigger for this arrangement was the disclosure by the Australian Prime Minister Anthony Albanese last week that OpenAI's model had infiltrated a government website in the country. More awkwardly, OpenAI later admitted that its AI had also accessed publicly available content from U.S. government websites, including pages from the Census Bureau and the Securities and Exchange Commission.
An OpenAI spokesperson confirmed on Monday local time that the company's Chief Strategy Officer Jason Kwon will fly from the United States to Australia next week to appear before a parliamentary committee for questioning. From providing funds to sending personnel, the company is trying to turn the trust crisis caused by "models acting on their own" into an opportunity for proactive security collaboration.
