17 April 2026

Welcome

I’m amused this week to see yet another call for some sort of “cyber reserve”, a sort of civil defence organisation to help out when we’re under cyber attack, because the idea has been around for ages.

I’m less amused to see that the tech trial into age assurance systems for social media ignored official privacy warnings.

And I am pleased in a comedy-neutral fashion by the wide range of topics this week. I hope you are too.

Oh, and thank you to the careful reader last week who noticed that I’d written “the war on Iraq” instead of on Iran. A correction has been made, but now I feel quite old.

Defence strategy has more AI, maybe less cyber?

Most of the 2026 National Defence Strategy and associated 2026 Integrated Investment Program released this week is outside the scope of this newsletter, apart from these notes:

• AI is now one of six priorities for innovation, wth defence minister Richard Marles saying AI holds the “most significant potential for technological disruption” in coming years. There’s further analysis at InnovationAus ($).

• Is cyber being downplayed? As Cyber Daily notes, while it’s getting more money, it isn’t one of the six pillars. The word “cyber” is used just 17 times in 102 pages, none of them in a headline.

For the record, the six pillars are: long-range fires and hypersonic weapons; high-energy lasers; autonomous systems; quantum technologies; artificial intelligence; and undersea warfare.

At The Conversation, strategic studies expert Peter Layton says the strategy feels written for a bygone era and ignores the elephant in the room: Australia’s reliance on the US as its major partner in an “increasingly unstable world”.

OAIC privacy warnings ignored by teen social media ban tech trial

The Office of the Australian Information Commissioner (OAIC) said that reporting from the tech trial for the teen social media ban was overstating the privacy protections, since the technologies hadn’t been assessed against privacy laws.

The OAIC repeatedly warned the trial team, but were repeatedly ignored, according to internal emails obtained via freedom of information by Cam Wilson at Crikey ($).

“The people behind the trial ignored most of those warnings and continued to include the inflated language in the final report, including terms like ‘privacy-preserving’ and ‘privacy by design”, Wilson writes.

“As scrutiny grows over how the teen social media ban is working in practice, these unaddressed concerns from an independent regulator cast doubt on the findings of the controversy-plagued trial used by the federal government to justify ploughing ahead with the social media minimum age policy.”

It’s almost as if the whole policy was rushed with unclear goals, and the tech trial was always intended to support the policy no matter what. At least that’s how it looks to me.

Ah yes that old Cyber Civilian Reserve idea...

Yet another call has been made for a cyber civilian reserve, something like the Army Reserve but for the cybers. Or, given its civilian nature, something like the State Emergency Service.

This time it’s come from Kersti Eesmaa, formerly Estonia’s ambassador to Australia but who now works on security-cleared personnel matters.

Estonia created a cyber reserve 15 years ago, Eesmaa told the Commercial Disco podcast.

“The lesson that we can learn here in Australia, from Estonia and other countries, is that you cannot be naive enough and think that you are protected, and that this will not happen [here],” she said.

“It takes time to create the proper reserve... and that’s why we have started [in Australia], so that we be ready if and when that would be needed.”

I wrote about this sort of thing a decade ago, and it wasn’t new then. A similar idea formed part of a cyber resilience program which was floated by Labor in 2020, when they were in opposition. Maybe this one will actually happen?

NEW PODCAST: If you read this newsletter, and it appears that you do, then may I suggest listening to The 9pm Cornucopia of Tech Policy Pleasures with Johanna Weaver and Zoe Jay Hawkins from the Tech Policy Design Institute in Canberra? In this episode we talk about how Australia as a middle power can participate in global tech policy. We chat about AI slop, the battle between Anthropic and the Pentagon, the digital duty of care, and of course the social media age restrictions. Just look for “The 9pm Edict” in your podcast app.

Also in the news

• From the Guardian, “The commonwealth ombudsman is investigating the government’s algorithm-based aged care assessment tool, which has been described by assessors as ‘cruel’ and ‘inhumane’ in its determination of home support funding for elderly Australians.”

• Australia is on the verge of an aged care AI boom, reports ABC News, but some experts have warned of the risks of unregulated AI, including “quite negative outcomes”.

• The Federal Court has published a Generative Artificial Intelligence Practice Note (GPN-AI), setting out the legal profession’s obligations when it comes to using AI in connection with court proceedings.

• Aussie youths are increasingly turning to AI for mental health advice.

• From The Conversation, “When AI starts shopping for you, fashion may be entering a new era of pricing.”

• Amazon has signed contracts with nine new renewable energy projects in NSW and Victoria, aiming to reach net zero carbon emissions by 2040.

• Digital technologies in rental housing favour landlords not tenants, according to a new paper by Samantha Floreani and Jathan Sadowski. That’s hardly surprising, but it’s good to have it said out loud in a peer-reviewed paper.

• “The first Australian to be prosecuted under Commonwealth laws introduced in 2024 to combat adult deepfake pornography has pleaded guilty,” reports ABC News.

• Federal opposition leader Angus Taylor says that the Coalition's immigration policy would introduce social media checks on all visa applications.

• “Apple has confirmed granting Australian law enforcement access to user notification data for the first time, which can identify a target’s device and potentially intercept or recover text from notifications,” reports Information Age.

• The latest report (PDF) from the National Anti-Scam Centre shows that in 2025 the total amount lost to scams was $2.18 billion. That’s up from $2.0 billion the previous year, but still lower than the peak of $3.1 billion in 2022.

• Latitude Finance was fined another $3.96 million for spam breaches, the second time they’ve been pinged.

• Roblox will launch restricted accounts for under-16s after being put on notice by the government two months ago.

• The Guardian has a profile of Noah Jones, the 15-year-old fighting the social media bans in the High Court. As it happens, he hasn’t been kicked off any of the platforms yet.

• NBN Co vastly overestimates the costs of providing better service, according to the ACCC.

• Coming up on 5 May, CSIRO Conversations: Quantum for every Business, being a free online thing. “Quantum is at the same stage AI was a decade ago, quietly moving from research labs into real world tools,” CSIRO writes, unironically — but seriously, I’m interested in seeing what’s actually coming out of those computers that look like chandeliers.

IF YOU’VE BEEN FINDING 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. And thanks to those of you who’ve already done so. Much appreciated.

Elsewhere

• The French government is migrating from Windows to Linux to reduce dependence on US tech, having already dumped Microsoft Teams and Zoom for video conferencing. The German state of Schleswig-Holstein had also dumped Microsoft, although it hasn’t been plain sailing.

• Seriously Risky Business reports on the idea that it’s time to ban the sale of precise geolocation.

• Fisherfolk found an underwater drone off Lombok. As H I Sutton wrote, it was probably a Chinese underwater sensor platform. But with no crew and no flag, what can Indonesia can legally do about it?

• The British Board of Film Classification is now using AI for censorship decisions. They rated all of HBO’s UK catalogues in six months, they say, whereas a human would’ve taken six years.

• One problem with AI systems is that they can fail without crashing. All the green lights are on, but users report that the system’s decisions are slowly becoming wrong.

• Scientists invented a fake disease called Bixonimania and uploaded two fake papers about it. Even faster than expected, the large language models (LLMs) sucked them in and AI told people it was real.

• The EU has announced it’s own age verification app, with EU-level privacy protections, although the bloc hasn’t decided what sites and services would require verification.

• Unlike many countries in Europe, Estonia is opposing social media bans for teens. Age-based bans are unenforceable because children will find ways around them, ministers argue. The government should invest in digital literacy rather than restricting young people’s participation in the information society.

• Google is promoting “teacher approved” apps for Android, although that doesn’t necessarily mean they’re educational.

• Here’s a fascinating history of encrypted messaging app Telegram’s complicated relationship with the Russian state, published by the Carnegie Endowment for International Peace last month.

• Also from last month, Oligarch Watch wrote about the billionaires who believe that humanity is just a warm-up act. How do you feel about some of the most powerful people in the world thinking of you as nothing more than “carbon‐based biological neural networks inside a cranium”?

• Meanwhile, Meta is building an AI replica of Mark Zuckerberg, reportedly a “photorealistic, AI-powered 3D” avatar, to converse with and offer feedback to employees. It’s amusing, sure, but there are indications that these AI chats could be used to make hiring and firing decisions. Will it have legs? Because...

• The inimitable Patrick Boyle’s latest video reminds us that Mark Zuckerberg spent $88 billion on a world with no legs. The money that Meta spent on the Metaverse is more in inflation-adjusted dollars than NASA’s entire Apollo program, and they had 10 legs on every mission!

• US-based lobby group Fight for the Future has launched a campaign to ban Meta Ray-Bans and similar technology from “family-friendly establishments” such as “your store, your place of worship, your school, your hospital, etc”.

• A nearly-defunct shoe company was about to be wound up, but then announced it was going to become an AI company. It’s share price rose 580%.

• There’s more on Iran’s information war machine in the New York Times (gift link).

• We mentioned last week that Amazon is killing off support for older Kindle devices. Well, it turns out you can jailbreak your Kindle and turn it into an open-source reader. I have not tried this.

• British satirist Michal Spicer voices the lead role in this delightful short film, Fan Art. Yes, it’s comedy, but it raises some interesting policy questions.

Inquiries of note

• Treasury is running an inquiry into the Census and Statistics Regulations 2026 remake. Submissions close 4 May.

• The Joint Committee of Public Accounts and Audit has kicked off an inquiry into the management of client privacy in the Australian public sector, and is currently accepting submissions. No timelines have been set.

Vale Gary Stark

Veteran Australian tech journalist and broadcaster Gary Stark died on Sunday 12 April aged 73 after years of declining health. His son Leigh has put together a beautiful memorial, and Influencing has published an obituary. The grumpy bastard will be missed.

What’s next?

Parliament is on break until Tuesday 12 May, when it’s Budget Night — but I’ll be watching out for public hearings for the various inquiries under way.

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.