The Weekly Cybers #82
Mumbling and grumblings about social media age restrictions because “It’s not a ban, it’s a delay”, Minister Ayers and the Productivity Commissioner talk up AI and talk down regulation, and much more.
22 August 2025
Welcome
Everything happens all at once! I have plenty of things to write about this week, from the discussions at TechLeaders to the Economic Reform Roundtable, but I’ve come down with the dreaded lurgi and my brain is all a'fuzzed.
That said, I’ve managed to churn out some words about two topics which have been dominating the news: the social media age restrictions, and the regulation of AI.
And of course there’s the usual burst of smaller stories — some of which are actually quite big so you should definitely click through.
eSafety Commission: “It’s not a ban. It’s a delay to having accounts”
In a delightful piece of linguistic sleight-of-hand, the eSafety Commission now says Australia isn’t implementing a social media “ban” for under-16s, it’s merely a “delay to having accounts”.
It’s a bit like kids not being banned from drinking alcohol. It’s just that their drink has been delayed.
“Age-restricted platforms won’t be allowed to let under-16s create or keep an account. That’s because being logged into an account increases the likelihood that they’ll be exposed to pressures and risks that can be hard to deal with,” eSafety wrote in a FAQ released earlier this month.
“Under-16s will still be able to see publicly available social media content that doesn’t require being logged into an account. As they won’t be logged in, they won’t be exposed to the more harmful design features of accounts.”
Those features include “the pressure to view disappearing content and respond to a stream of notifications and alerts”, which eSafety says have been linked to health harms such as “reduced sleep and attention and increased stress levels”.
Commissioner Julie Inman Grant first used this framing of a “delay” in a speech in June.
And while we’re on the topic of social media...
- Key stakeholders in the age assurance tech trial have been frozen out amid media leaks and resignations, and the advisory board may not see the final report until the communications minister releases it in the coming weeks.
- The social media ban — sorry, “delay” — won’t have “critical” safety measures in place for its 10 December kick-off, thanks to the government’s self-imposed deadline. “Department of Communications staff said important recommendations would only be considered in a 2027 review of the law.”
- From The Conversation, “The social media ban is coming, whether families like it or not: 5 ways to prepare kids and teens”.
Minister Ayers’ incuriously bland faith in AI and his odd framing of the media’s role
Senator Tim Ayers, Minister for Industry and Innovation and Minister for Science, gave a distinctly uninspiring speech about artificial intelligence at TechLeaders on Monday. Some hand-waving about AI’s “unmissable opportunity”, sure, but no specifics.
“If we all work together to adopt early, invest strategically, and give workers, businesses, managers and researchers the capacity to use AI effectively, we will ensure that the benefits of AI accrue to everyone, not just some,” he vagued without any apparent enthusiasm.
And then he said this:
“It is incumbent on all of us — in government, in business, in the union movement, in civil society groups and think tanks, and especially in the media — to build the confidence of Australians as we work through all the questions that AI brings up.”
No, Minister. It is most definitely not the role of the media to “build the confidence of Australians” — especially if confidence is not warranted.
The media should be looking at the AI industry’s upbeat claims with a critical eye and report the technology’s strength and flaws so we can all make informed decisions.
A government minister who swallows the tech industry’s hype wholesale, and expects others to do the same, is a waste of carbon atoms.
Productivity Commissioner’s upbeat AI deregulation push
A “new and overarching regulatory framework” for artificial intelligence is not the way to go, according to Productivity Commissioner Danielle Wood.
“The risks posed by AI are mainly existing risks. AI may make it cheaper, easier and faster for bad actors to create harms, but most of these harms — from product safety, to discrimination, to fraud — are already covered by regulatory frameworks,” Wood told the National Press Club on Monday.
“The process of identifying any ‘gaps’ in these frameworks and finding amendments to address them is quietly going on throughout government.”
The Productivity Commission estimates that AI is “likely” to add more than 4% to labour productivity over the next decade.
The commission has previously called for copyright reform to allow AI models to use the copyrighted material of Australians without permission as “fair dealing”.
However the commissioner “errs” in “assuming that removing regulatory impediments alone will unlock its benefits for Australia”, according to Dr Marina Yue Zhang from the Australia-China Relations Institute at UTS.
“History shows that the diffusion of such technologies does not occur automatically; it requires deliberate strategy, patient capital, and public–private collaboration,” Zhang writes.
“Australia does not need to mimic the United States or China. But it must resist the illusion that ‘getting out of the way’ will suffice... If AI is to be Australia’s electricity moment, then policymakers must do more than deregulate,” she said.
“They must lay the wires, fund the research, and shape the rules that will determine whether the benefits flow widely — or bypass us altogether.”
STOP PRESS (or whatever the digital equivalent is): The Australian Council of Trade Unions (ACTU) claims that it’s reached a “breakthrough” agreement with the tech sector — by which they mean the big-tech lobby group Tech Council of Australia — on a model for paying to using creative content in training AI. Groups representing the creative industries have dismissed those claims. I haven’t head the headspace to go through this yet. I’ll have more next week.
Also in the news
- Australian live-streaming platform Kick broadcast a man’s death, so that’s something to think about.
- In Australia’s latest big fat data breach, the email addresses and phone numbers of around 280,000 iiNet customers have been exposed, along with 10,000-odd iiNet user names and street addresses, and about 1,700 modem set-up passwords.
- From AAP, “Most Australian workers think artificial intelligence will impact their jobs but few believe it will replace them entirely, and many remain curious about what it can achieve.”
- Apparently many Australians secretly use AI at work. They hide it because they “feel that using AI is cheating”, have a “fear of being seen as lazy”, and a “fear of being seen as less competent”. Clearer rules could reduce this so-called “shadow AI”.
- A WA lawyer has been referred to the regulator after submitting AI-generated documents citing nonexistent cases.
- The Australian Securities and Investment Commission (ASIC) is expanding its scam website takedown capability to include ads on social media. They’ve also listed the top five online investment scam trends.
- Meanwhile CommBank has deployed AI bots to keep scammers busy on voice calls. The bank has also apologised for the “error” of replacing workers with AI as call volumes rose.
- Google has agreed to pay $55 million in fines for its past agreement with Telstra and Optus to pre-install only Google Search in their Android phones in exchange for a cut of the advertising revenue.
- The Australian Prudential Regulation Authority (APRA) warns there’ll be more cyber attacks on financial institutions thanks to increased geopolitical tensions.
- The brainwaves from Tech Council chair Scott Farquhar have featured last week and the week before, but I want to come back to another drawback with his suggestion that Australia could become a regional data centre hub: How are you going to keep it cool given our limited water supplies? Because data centres use a lot of water.
- Home Affairs is running a survey as part of its Impact Study for Small and Medium Enterprises regarding Cyber Security Supports. It’s basically asking which of their projects you’ve heard of and such, rather than what your needs might be, but it’s probably worth giving them some feedback.
NEED A MUSIC BREAK? Why not try the second episode (not counting the pilots) 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.
Elsewhere
- Agentic AI web browsers, which act autonomously on users’ behalf, are extremely gullible and unsafe to use because they fall for scams and phishing attacks.
- The UK has dropped its demand for a backdoor into Apple encryption.
- “At least six prominent publications have published AI-generated articles by a fictitious freelance author named Margaux Blanchard over the past few months,” reports Mumbrella.
- Facial recognition software vendors make some impressive claims about their accuracy, but they appear to be significantly less accurate in real-world settings.
- AI free from bias and ideology is a fantasy because humans can’t organise data without distorting reality.
- Thanks to TikTok, they reckon, the average length of a pop song is down to 3 minutes 12 seconds, but correlation is not causation.
- A brilliant headline at Defector: “It took many years and billions of dollars, but Microsoft finally invented a calculator that is wrong sometimes”. Also, good point.
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.
Inquiries of note
Nothing new this week.
What’s next?
Parliament returns this coming Monday 25 August.
The Senate draft legislation program shows nothing of interest to this newsletter, but of course new things are often added on the day.
In the House of Reps debate resumes on the Treasury Laws Amendment (Payments System Modernisation) Bill 2025.
DOES SOMETHING IN THE EMAIL LOOK WRONG? Let me know. 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.