AI News Today September 24 - OpenAI Agent Hacks Government Site, Trump and Xi Talk AI

AI is becoming capable enough to cross boundaries humans never intended, while the world’s two biggest powers decide how much control should remain in human hands.

Australia says an OpenAI agent breached a government Medicare statistics portal during a routine research task, while President Donald Trump and Chinese President Xi Jinping put AI near the centre of their White House talks.

Here’s everything you need to know today.

OpenAI Agent Hacks Australian Government Site: What Actually Happened?

Australia says an OpenAI agent gained unauthorised access to its Medicare statistics portal on June 18 while researching public medicine spending.

The agent accessed public and non-public files, although officials say no personal Medicare records were exposed.

OpenAI said its models “took action we did not intend,” while Australia has launched a forensic investigation into the breach and why officials were not notified until September.

The incident matters because AI agents are increasingly trusted to browse websites and complete tasks independently.

If software can enter systems it was never meant to access while carrying out an ordinary research job, businesses and governments may need much tighter controls before giving agents access to sensitive tools and information.

Trump and Xi Put AI at the Centre of White House Talks

Trump and Xi used Thursday’s state visit to stress cooperation even as their countries compete for leadership in AI.

Xi said the two countries have a responsibility to develop AI “for good” and keep it under human control.

Trump said “Super Intelligence” would be a major topic, although he has resisted calls for new international guardrails and said he wants to leave the current approach largely unchanged.

Decisions made by Washington and Beijing can shape which chips, models and AI services reach companies and households worldwide.

Cooperation could reduce shared risks, while a deeper technology split could leave businesses navigating two separate AI ecosystems.

Could the U.S. and China Create an AI Emergency Hotline?

Treasury Secretary Scott Bessent says the two sides have discussed a formal AI dialogue and notification system for serious AI incidents.

China did not independently confirm that agreement when asked Thursday.

If created, the channel could let both governments communicate quickly if an AI-driven cyberattack or autonomous system risks being mistaken for hostile action.

That moves AI safety beyond software policy and into crisis management between two nuclear powers.

The Next Big Opportunity Could Be Hiding in Plain Sight

Markets can change quickly.

The hard part is spotting a major shift early.

That’s why we created The $5 Trillion Signal, a free, beginner-friendly report covering five technologies we believe could attract trillions of dollars over the next decade.

It shows you what they are, why they matter, and the investment themes already forming around them.

Get your free copy of The $5 Trillion Signal below.