The Weekly Cybers #86
Four people die as Optus breaks Triple Zero calls, Australia’s looming social media age restrictions continue to raise questions, OpenAI’s ChatGPT features in a raft of AI stories, and more.
26 September 2025
Welcome
Optus just seems to lurch from crisis to crisis. In the same week that the Federal Court confirms a $10 million fine for “unconscionable conduct”, the telco fesses up to a major Triple Zero systems outage which has been linked to four deaths.
Meanwhile Australia’s looming social media age restrictions remain in the news, with doubts and uncertainties failing to confine themselves to rigidly defined areas.
You may have noticed that I reference the Guardian and Crikey quite a bit in these stories. That’s because they cover digital policy issues in some detail. They’re not alone, of course. ABC News does a credible job, as do the various IT news operations. But I don’t have time to read everything, and a bias my choices to non-paywalled or leaky-paywalled sites.
And as usual, there’s a lot of stories relating to AI. OpenAI’s ChatGPT is high on the list this week.
Four die as Optus breaks Triple Zero calls
It’s a big one. As The Saturday Paper’s Post newsletter summarised, “An Optus outage last week has been linked to four deaths, and pressure is mounting for the telco to be penalised”.
Apparently a botched firewall update, described as “regular” maintenance at 12.30am last Thursday, caused Triple Zero calls on the network to fail for 13 hours.
As iTnews reported, “[Optus CEO Stephen Rue] said that the telco’s telemetry did not show anything abnormal after the ‘regular’ change was made, although he did note that the carriage of emergency calls is not specifically monitored”.
During the incident there were four deaths when Triple Zero calls failed: an eight-week-old boy and a 68-year-old woman in Adelaide, and a 74-year-old man and a 49-year-old man in Perth, although SA Police says the outage was “unlikely” to have caused the baby’s death. I’m sure that consoles the parents.
Optus says the failure was down to human error. Your writer is of the view that there’s no such thing. It’s the system that has failed when it doesn’t protect the human from making an error.
Mistakes do happen, of course, but Optus has seen quite a few mistakes over the past couple of years.
There’s a lot more going here, so here’s a quick summary.
- Five months ago Optus said it would be a “huge burden” to provide real-time updates on emergency call outages.
- From AP, “Australia will overhaul telecommunications after Optus emergency call failure linked to 4 deaths”.
- For more than a year now, there’s been talk of and even a promise to create a triple zero custodian, but it hasn’t happened yet. How would that help?
- Via The Conversation, “The Optus brand is in tatters. How can it even begin to rebuild customers’ trust?”
- Also from The Conversation, “Should the Optus chief quit? These 5 fixes would do far more to stop another 000 failure”.
Court confirms Optus’s $100,000 “unconscionable conduct” penalty
Meanwhile, Optus has been ordered to pay a penalty of $100 million by the Federal Court this week.
The court was confirming a agreement made between Optus and the ACCC after the telco admitted to engaging in “unconscionable conduct” when selling mobile phones and contracts to hundreds of Australians, many of whom could not use or could not afford what they were sold.
“Many of the affected consumers were vulnerable or experiencing disadvantage, such as living with a mental disability, diminished cognitive capacity or learning difficulties, being financially dependent or unemployed, having limited financial literacy or English not being a first language,” the ACCC wrote.
“Many of the consumers were First Nations Australians from regional, remote and very remote parts of Australia.”
In some cases debt collection actions had continued even though Optus was aware that the debts had been entered into at Optus’s Mount Isa store “fraudulently and without consumer knowledge”.
More discourse on social media age restrictions
As 10 December slowly approaches, we can expect to see more discussion of Australia’s social media age restrictions, especially as we see more fallout from the UK’s version which is already in force.
Here’s what I noticed this week:
- Elon Musk’s X has called for a delay in the start date, citing “serious concerns” about the policy’s lawfulness under human rights treaties.
- Instagram will be using AI to
determineguess users’ ages. Meanwhile a whistleblower says the platform still poses a risk to children.
- A Guardian test discovered that even when logged out, scrolling shortform videos on YouTube and TikTok quickly leads to problem content including gambling ads, and violent and far-right content.
- Half of Australians have no idea that age checks are coming to search engines. Interestingly, while 46% of respondents said they were very comfortable or comfortable with social platforms checking their age, when they were presented with the different ways that it could be done, no single method was considered comfortable by a majority.
- Age restrictions could be extended to Reddit, Twitch, Roblox, Lego Play, gaming platform Steam, and even dating apps.
- The eSafety Commissioner has appointed an advisory group “to support a robust and transparent evaluation of the implementation and outcomes of the Social Media Minimum Age (SMMA) obligation”, led by members of the Stanford University Social Media Lab.
- European Commission president Ursula von der Leyen praised Australia’s move as "plain common sense".
- From The Conversation, a public health perspective: “Banning something to reduce harm is a tricky business. Here’s what works.”
- And from last week, The Register has further analysis of the regulatory guidance issued by the eSafety Commissioner.
IF YOU FIND 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. Please consider.
Also in the news
- The Australian Federal Police (AFP) has been using encrypted messaging app Signal, according to emails obtained by Crikey in a freedom of information request. An AFP spokesperson said that personnel were expected to comply with record-keeping rules — which is not the same as saying that they are.
- NBN Co says it’s rolling out fibre upgrades and backup gear as it prepares for more frequent extreme weather events.
- Australia may have to choose between a Chinese TikTok and an American one owned by a consortium of Trump’s billionaire backers, following his signing of an executive order supporting the deal. China has yet to approve the deal.
- Former SBS chair George Savvides says online TV networks should follow SBS’s lead with opt-out feature to protect people from gambling ads and other “addictive and destructive” advertising. Your humble writer suggests we should be able to opt-out of any category we like.
- The Digital Transformation Agency (DTA) has launched a new platform to track government IT spending. According to iTnews it’s a combination of a CRM and a case management system.
- The AI bubble looks like it’s about the burst, and that could be good for Australia, reports ABC News. But earlier in the week they reported that AI will be the biggest gold rush as it revolutionises the Australian economy.
- An AI chatbot encouraged an Australian man to murder his father, although I believe he didn’t actually do it.
LATEST PODCAST: I chat about conspiracy theories in Australia in The 9pm Conspiracy Nation with Ariel Bogle and Cam Wilson, authors of the book titled, you guess it, Conspiracy Nation. Look for “The 9pm Edict” in your podcast app of choice.
Elsewhere
- Reuters has published some astounding journalism on what it’s like to be kidnapped and taken to a scam camp in Myanmar. Not pleasant reading.
- Via Cyber Daily, “‘Death is not the end’: US megachurches play AI-generated clip of Charlie Kirk addressing his death,” and it scored a standing ovation.
- In the US a judge has granted preliminary approval to AI company Anthropic’s deal to compensate authors with US$1.5 billion. It sounds a lot, but with seven million books in the pirated database that works out at a mere US$214 per book.
- The Internet Archive’s big battle with music publishers has ended in a settlement. The battle was over the Great 78 Project, an effort to preserve early music recordings that only exist on brittle shellac records.
- Facebook data has revealed misinformation’s real-world impact, including poor health outcomes, falling public trust, and significant societal harm.
- OpenAI has kicked off an experiment in ChatGPT which proactively does research to deliver personalised updates based on your chats, feedback, and connected apps like your calendar”. So, yet another grab for more eyeball time.
- On the plus side, OpenAI boss Sam Altman says ChatGPT will stop discussing suicide with kids.
- And from last week, “OpenAI admits AI hallucinations are mathematically inevitable, not just engineering flaws”.
Inquiries of note
- The Parliamentary Joint Committee on Law Enforcement (PJCLE) is inquiring into Combatting Crime as a Service. Submissions close 13 October. This actually kicked off earlier this month but somehow I missed it.
- The Australian Communications and Media Authority (ACMA) has proposed allowing people buying pre-paid SIMs to use digital ID rather than having to produce hard copies. Submissions close 21 October.
- Treasury has released exposure-draft legislation for the so-called “crypto bill”, aiming to regulate digital asset platforms. Breaches could cop the culprits fines of $1.6 million or more. Submissions close 24 October.
- The Responsible AI in Marketing Council is seeking industry feedback on its draft Industry Guidelines on Ethical AI Use in Marketing. Submissions close 31 October.
- The Digital Platform Regulators Forum (DP-REG) has published a working paper on immersive technologies, by which they mean virtual reality (VR), augmented reality (AR) and such. They’re seeking feedback but no deadline was given.
What’s next?
Parliament is currently on a break until 7 October, when the House of Representatives returns and Senate Estimates hearings are held. That’s 11 days from now.
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 weekly digest of 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 a cyber security newsletter. For that that I recommend Risky Biz News and Cyber Daily, among others.