CAMERON WILSON

NEW: A man asked his AI assistant OpenClaw to book a gym class. It found a exploit in the gym website, got around booking restrictions and kicked someone off the waiting list to move him up a spot It's the first known Australian case of AI agents autonomously hacking! www.abc.…

How a simple request for AI to book a gym class exposed a major threat abc.net.au

NEW: OpenAI's agents also tried unsuccessfully to hack another Australian government agency, say researchers who believe this is "first reported instance of agents hacking a government". They claim bots left traces of trying to exploit a vulnerability on the AIHW website www.a…

How OpenAI agents tried to thwart cybersecurity amid Medicare hack abc.net.au

NEW: AI music is going to be banned from Australia's top music charts unless it can prove that it was "substantially human-made" This comes weeks after a song with AI vocals became Australia's most played song www.abc.net.au/news/...

AI-generated songs banned from Australian music charts abc.net.au

NEW: ABC staff have been told there's no question whether the broadcaster will use AI, as it rolls out new tools, hires AI specialists and trials using AI to help write web articles from radio news. Experts & the MEAA cautiously welcome plan but warn misuse could hurt trust ww…

ABC trials AI writing tools for news staff amid trust warnings - ABC News abc.net.au

NEW: Dozens of AI data centre projects will escape Australia's proposed energy & water restrictions which aren't expected apply to green-lit projects. While the Prime Minister said they're "smaller", these projects are generally bigger and power hungrier www.abc.net.au/news/...

Bigger and hungrier: What we know about 27 proposed new data centres abc.net.au

AI has just had its Sarah Connor moment. Last week's rogue AI story has been dismissed by some as marketing. But even a cynical read of the news & other evidence suggests it should be a "warning shot" for the serious disruptive threat of AI that will soon be widely available w…

AI just had its Sarah Connor moment. Is Australia ready? abc.net.au

Data centres powering AI could demand as much new electricity as cars & houses until 2040, per Greenpeace analysis. It's not just the size of the energy demand, but also its immediacy: forecasts for data centre energy use jumped 6x over the last year www.abc.net.au/news/...

'None of this was forecast': Energy demand from AI data centres soars abc.net.au

Scoop: Australia has picked a philosopher and air force reservist to lead a new body tasked with keeping Australians safe from the harms of AI. Dr Kate Conroy will lead Australia's AI Safety Institute, which will monitor, test and share info on emerging threats from AI. www.ab…

Philosopher tasked with keeping Australians safe from AI abc.net.au

NEW: Dozens of Australian government agencies missed their first test of self-policing their AI use. While these agencies are now compliant, the lapses show pitfalls in Australia's decision to task existing regulators to make their own AI rules instead of a central AI act. www…

Government agencies fail first hurdle under AI self-reporting policy abc.net.au

OpenAI put out research this week which found no correlation between revenue per employee and use of AI, Fortune reports.

re: OpenAI Medicare data hack, the intrusion wasn't sophisticated nor did it access particularly sensitive data as far we know. What is noteworthy is how OpenAI's models did it without being explicitly asked. The fear is what happens if this tech get more capable and accessible

Getting to know the AI beat, one thing has come up with ordinary people, industry, politicians, academics, activists: Most Australians do not trust AI or the people behind it. It's a big issue for an industry that needs people to adopt its products and share their resources.

a truly unfortunate chryon to have under an interview about AI

Here is the email obtained by the ABC that was sent by OpenAI to the Australian government's Services Australia to disclose the incident

An OpenAI spokesperson confirms that its models did access internal Australian government files, but say that it was only "aggregate health statistics and internal file names." They say the hack happened in June, was discovered in August, and they reported it Sept 10

3 days before OpenAI emailed the Australian government to disclose the hack, I wrote about the need for mandatory AI incident reporting. It's hard to think of any event that would jeopardise social license more than finding out an AI company covered up the breakout of a rogue AI

Prime Minister Anthony Albanese says OpenAI's "AI agent infiltrated" a Australian government website for healthcare statistics, accessing non-public information. PM says the company took three months to tell Australia.

NEW: Leaked documents shows the Albanese government has floated 2 ways to let AI companies train their models on Australian content One would let companies like OpenAI and Anthropic do this without paying some creators if they strike enough deals deals with enough rights holders

OpenAI has apologised for hacking Australia’s Medicare statistics system in a blog post that revealed the intrusion was broader than first known.