- OpenAI reportedly used an LLM to help draft and edit its Australian government breach notification, though humans reviewed and sent the final email.
- The revelation followed Jason Kwon telling a parliamentary hearing that he did not think AI had been used to prepare the notification.
- Assistant minister Andrew Charlton argued that AI requires regulation because industry self-regulation cannot adequately address its risks and incentives.
- Former RBA official Jonathan Kearns warned AI investment could add near-term inflationary pressure and contribute to further interest-rate rises.
The saga around OpenAI and the Australian government continues. The Guardian Australia reported that OpenAI used an LLM (large language model) to draft the email it sent to the Australian government about the unauthorised access to several government sites.
This comes just days after chief strategy officer Jason Kwon said at a parliamentary hearing he didn’t think that AI had been used, but he’d check.
The Guardian cited a person with “knowledge of the incident” who said that while AI was used in part to draft and edit the email, the final review and actual sending were handled by humans. OpenAI had not responded to requests for a comment.
In earlier reporting, the outlet also took aim at Kwon’s statements before senators, saying his responses sounded as though they had been written by a chatbot.
The Guardian said his polite tone and agreeable manner sounded much like the way a chatbot replies to a user’s query.
As previously reported, several incidents occurred over the past few months in which Australian government websites were accessed. At times, OpenAI was slow to report the incidents. Kwon acknowledged this; he also promised to do better.
Read also: Sui Tunnels Smash 40 Million TPS in AI-Agent Stress Test
Australian AI Regulation
When it comes to AI regulation, Australia is going with a different approach from the US. In the States, the current way is to let industry regulate itself. Assistant minister for science and technology Andrew Charlton said such an approach wasn’t workable and that “the market will not fix” problems arising from AI development.
AI needs regulating because its harms are severe, hard to undo, borne by people who never chose them, and sometimes invisible until they arrive.
Andrew Charlton, assistant minister for science and technology He added that self-regulation would fail “because the incentives reward speed and capability, and even the people at the top of the industry cannot slow down by themselves”.
AI Causes Inflationary Pressure, Former RBA Staffer Says
Meanwhile, former Reserve Bank of Australia (RBA) official Jonathan Kearns warned that AI investment is likely to cause short-term inflation. Speaking at the Reuters Global Markets Forum, he said that it would take time for productivity benefits to show, while investment in technology and infrastructure would put pressure on domestic prices, ultimately leading to rate hikes.
Read also: Ripple’s TradFi Ambitions Grow with Wall Street Financing and Korea Deal
The RBA recently raised the cash rate by 25 basis points, taking it to its highest level since 2011. Kearns expects another rate hike this year.
There’s absolutely no chance [the RBA] can look through the inflationary impetus because we’ve had four or five years of high inflation and inflation being above target.
Jonathan Kearns, former Reserve Bank of Australia (RBA) official AI,OpenAI#OpenAI #Draft #Australian #Government #Breach #Notification1791533510
