The Weekly Cybers #80
The Productivity Commission and the Treasurer side with tech bros on AI regulation, NBN goes with Amazon’s Project Kuiper, OpenAI launches GPT-5, and much more.
8 August 2025
Welcome
Whew! There’s nothing about social media age restrictions this week. Well, apart from a brief note about what’s happening in the UK. No, this week it’s all about the Productivity Commission proposing that AI companies should be able to let rip on copyrighted material.
Not everyone likes that idea.
There’s also the usual raft of smaller stories — smaller in that I’m just lining to other people’s reporting — but there’s some fascinating stuff happening.
And how many Bs are there in “blueberry”? The answer might surprise you.
Productivity Commission, Treasurer bullish on AI
“The Productivity Commission [PC] is examining whether technology firms should be exempt from copyright rules that [might] stop companies mining text and data to train artificial intelligence models,” reports the Guardian.
This was just one of a bunch of ideas in the commissions’s interim report on harnessing data and digital technology released this week.
“[One] option is to expand the existing ‘fair dealing’ regime, which provides certain exceptions to the requirement to obtain permission from the copyright holder... Currently, there is no exception that covers AI model training per se... However, depending on the case, a different exception could apply. For example, AI models built as part of research could fall within the scope of the ‘research or study’ exception.”
The PC’s report also floats the idea of delaying the introduction of any “mandatory guardrails for high-risk AI”, and dumping the idea of an EU-style right to erasure because “this would impose a high compliance burden on regulated entities, with uncertain privacy benefits for individuals”.
In essence, the PC reckons that regulation would get in the way of people making money.
Productivity Commission sides with the tech bros
The PC’s ideas echo the thoughts of Scott Farquhar, co-founder of must-mention tech company Atlassian and, more importantly, chair of the big-tech Tech Council of Australia.
“Australia’s copyright laws are out of sync with the rest of the world,” Farquhar said in his speech at the National Press Club last week.
“Today, large language model providers don’t want to train their models in Australia... We are in a perverse situation where copyright holders aren’t seeing any more money, but we also don’t see the economic upside of training models in Australia.”
Aside: Last week I suggested that Scott Farquhar’s proposal to make Australia a data centre hub for South-East Asia seemed odd, because network latency would be a problem. I’ve since been told about a firm in Singapore keeping their data in Australia, and “No one has complained about latency. Yet.” Maybe I talk with too many gamers.
Forbes has more on Farquhar’s five-point AI plan.
As Josh Taylor writes at the Guardian, the PC appears to have bought into tech companies’ brazen arguments.
Indeed, Treasurer Jim Chalmers is openly siding with the PC.
“The PC’s broad directions are largely consistent with the directions that I set out on the weekend,” Chalmers said. AI should be treated as “an enabler, not an enemy”.
That said, industry minister Tim Ayes says there are no immediate plans ($) to change copyright laws — which of course doesn’t mean “immediate plans” might not appear tomorrow.
PC proposals at odds with expert’s AI ethics framework
Meanwhile, one of the world’s leading AI experts, Professor Toby Walsh, has outlined his framework for ethical AI, based in part on the ethical framework used in medicine: the Hippocratic oath.
Its four pillars are beneficence, non-maleficence, autonomy, and justice.
As The Mandarin reports, Walsh proposes that the “seeming unpredictability” of AI warrants a fifth precautionary principle.
“We are going to have to think carefully about the long-term consequences of things,” he said.
“Lots of people are starting to use AI bots as therapists or companions. I think we need to be rather careful about the consequences of that.”
Written submissions addressing the Productivity Commission’s interim report close at 5pm on AEST Monday 15 September.
NEED A MUSIC BREAK? There’s a new episode of Another Untitled Music Podcast, in which my good friend Snarky Platypus and I bring you a selection of songs with “Stay” in the title. Look for it in your podcast app of choice.
Also in the news
- The Commonwealth Ombudsman has found that 964 jobseekers had their income support payments unlawfully terminated by The Machines. The Department of Employment and Workplace Relations had failed to implement the post-robodebt legislation which was passed two years previously.
- The Australian Information Commissioner (AIC) is launching a civil penalty action against Optus over it’s 2022 data breach. “Optus seriously interfered with the privacy of approximately 9.5 million Australians,” claims the AIC, so with a potential fine of up to $2.22 million for each contravention, that could come to $21.09 trillion. But of course it won’t.
- NBN Co has signed onto Amazon’s Project Kuiper as a replacement for its SkyMuster satellite internet service, rather than SpaceX’s Starlink. The risk management of this choice presents an interesting dilemma: sign onto a system which is not yet operational, of go with the working system run by the volatile Elon Musk?
- The eSafety Commissioner has slammed the tech giants for leaving “significant gaps” in their efforts to combat certain types of problematic material which I won’t mention here because it caused some email providers to block last week’s edition.
- An interesting argument via The Conversation: “If you have decades of work experience, your skills, knowledge, and judgement could be exactly what’s needed to harness AI’s power — without falling into its traps.”
- From ABC News, “Health practitioners are increasingly using generative AI to help them record and summarise patient information. Here is what to know before you agree to them using it.”
- The Australian Defence Force has been advertising on TikTok and in video games, resulting in a 15-year high in recruitment numbers.
- An Australian man using a false name — presumably to avoid people realising he has a “colourful past” — has raised $19 million from investors for an “ethical internet” but nothing has yet come of it. Technically it’s not a dodgy share scheme because he’s not selling shares as such but... oh just read it for yourself.
IF YOU’VE FOUND THIS NEWSLETTER HELPFUL, PLEASE SUPPORT IT: The Weekly Cybers is currently unfunded. It’d be lovely if you threw a few dollars into the tip jar at stilgherrian.com/tip.
Elsewhere
- OpenAI has launched GPT-5, the latest version of its AI model. It thinks that the word “blueberry” has five Bs. Yes, that’s been confirmed. But there is some good news: it’s more sarcastic.
- In the UK, all manner of content is being age-gated under their Online Safety Act, including “Reddit communities about stopping drinking and smoking, periods, craft beers, and sexual assault support, not to mention documentation of war”. Meanwhile VPN use has exploded.
Inquiries of note
Nothing new this week, apart from the Productivity Commission’s biggie.
Related news digests
I’ve mentioned it before, but the free newsletter The Crux from tech analysis and consulting firm PivotNine is one of my must-reads every Tuesday. It primarily covers enterprise technology, cloud, and cybersecurity, but Justin Warren’s observations on the state of the technology industry are well worth reading.
What’s next?
Parliament returns on 25 August, which is 17 days away.
On Wednesday 13 August the Senate Legal and Constitutional Affairs Legislation Committee is holding a public hearing on the Whistleblower Protection Authority Bill 2025.
DOES SOMETHING IN THE EMAIL LOOK WRONG? If there’s ever a factual error, editing mistake, or confusing typo, it’ll be corrected in the web archives.
The Weekly Cybers is a personal look at what the Australian government has been saying and doing in the digital and cyber realms, on various adjacent topics, and whatever else interests me, Stilgherrian, published every Friday afternoon (nearly).
If I’ve missed anything, or if there’s any specific items you’d like me to follow, please let me know.
If you find this newsletter useful, please consider throwing a tip into the tip jar.
This is not specifically a cyber security newsletter. For that that I recommend Risky Biz News and Cyber Daily, among others.