The Weekly Cybers #91
AI companies cop a copyright slap down, Microsoft cops a legal battle over misleading conduct, and cops starting building an AI-based Gen Z and Alpha translator. Oh, and lots more.
31 October 2025
Welcome
There’s some big stories this week. The Australian government slapped down AI companies’ demands for free content. Microsoft is being taken to court over alleged misleading conduct. And if you used Facebook a decade ago, you might be eligible for compensation.
While there is, as usual, a bunch of stories involving AI, there’s further evidence that the bubble may be bursting soon. Nvidia has become the world’s first $5 trillion company, but so much of this value is based on circular financing. No one is actually making any real money here — except perhaps the energy companies.
And also as usual, there’s much more — including an overhyped humanoid robot.
AI scrapers won’t get free rein in Australia
The Productivity Commission had wanted the government to grant a copyright exemption to allow tech companies free access to train their AI models, but the government said no.
“The government stands behind Australia’s creative industries and, by ruling out a Text and Data Mining Exception, is providing certainty to Australian creators,” said Attorney-General Michelle Rowland in a press release.
The government convened its Copyright and AI Reference Group (CAIRG) to discuss three priority areas: whether a new paid collective licensing framework under the Copyright Act should be established for AI usage, or some voluntary scheme; “explore opportunities to clarify or update how copyright law applies to material generated through the use of AI”; and “make it easier to enforce existing rights through a potential new small claims forum to efficiently address lower-value copyright infringement matters”.
“Artificial Intelligence is an exciting technological frontier full of opportunities,” Rowland enthused, meaninglessly.
So in brief, we don’t know exactly what will be happening, but it won’t be giving AI companies free rein.
Microsoft misled consumers, ACCC alleges
Microsoft is being taken to court by the Australian Competition and Consumer Commission (ACCC) for allegedly misleading around 2.7 million Australians over their Microsoft 365 subscriptions.
“The ACCC alleges that since 31 October 2024, Microsoft has told subscribers of Microsoft 365 Personal and Family plans with auto-renewal enabled that to maintain their subscription they must accept the integration of [the AI capability] Copilot and pay higher prices for their plan, or, alternatively, cancel their subscription,” the commission wrote in a press release, which also included the relevant screenshots.
“Microsoft’s communication with subscribers did not refer to the existence of the ‘Classic’ plans, and the only way subscribers could access them was to begin the process of cancelling their subscription.”
However as PR practitioner Phoebe Netto writes at Mumbrella, Microsoft knows that you need them more than they need your trust. “A breach of trust that would bring most other businesses to their knees is unlikely to put a significant dent in Microsoft’s sales,” she writes. She’s not wrong.
As an aside I have seen — but can’t find just now — social media posts claiming that Microsoft’s corporate salespeople’s main KPI this quarter is Copilot sales. If you can confirm or debunk that, I’d like to hear from you.
AFP is building an AI-based Gen Z translator
The Australian Federal Police (AFP) is warning of groups of boys and young men, overwhelmingly from English-speaking backgrounds, who are grooming female victims and then forcing them to “perform serious acts of violence on themselves, their siblings, others or their pets”.
In her speech to the National Press Club, newly-appointed AFP Commissioner Krissy Barrett said these “crimefluencers” — her word — are “motivated by anarchy and hurting others”.
“To be accepted into these networks, the perpetrators often have to pass a test or undertake a task, such as providing videos of the self-harm of others, or other gory content,” she said.
“In this new, twisted type of gamification, perpetrators reach a status or new level in their group when they provide more content showing more extreme acts of depravity and sadism.”
AFP is working with Microsoft to develop “a prototype AI tool that will interpret emojis and Gen Z and Alpha slang in encrypted communications and chat groups to identify sadistic online exploitation”.
LATEST PODCAST: Last week Journalist Erin Cook and I discussed Philippines politics and other South-East Asian issues in The 9pm Bongbong in the Region with Erin Cook. Look for “The 9pm Edict” in your podcast app.
Also in the news
- Were you on Facebook 10 years ago? You may be eligible for compensation. Some 311,000 Australian Facebook users had their privacy breached and there’s $50 million up for grabs.
- Following complaints about the new Bureau of Meteorology (BoM) website last week, environment minister Murray Watt has asked for “urgent consideration of improvements to the website’s functionality and usability”, among other things. The Bureau has apologised, but yeah (some) people are not happy. Meanwhile The Mandarin asks, is it really a disaster?.
- “Most Australian government agencies aren’t transparent about how they use AI,” reports The Conversation.
- “Home Affairs is streamlining its security vetting procedures for Commonwealth technology suppliers,” reports iTnews.
- And while we’re at iTnews, “The federal government will go ahead with a plan to direct telcos to establish a public register of network outages”. I wonder how fine-grained it will be, because random little outages affecting small numbers of customers are happening all the time.
- The Australian Strategic Policy Institute has some analysts of AI and the future of espionage.
- Australia’s financial crimes regulator AUSTRAC has taken action against a second cryptocurrency ATM operator, Cryptolink.
- Australian AI company Maincode has announced a $30 million investment to build an advanced AI factory in Melbourne, due to open in January 2026.
- From The Conversation, “When you click on an ad in sales season, retailers get to harvest your data”. This happens everywhere and everywhere, of course, but it’s good to get a reminder from time to time to time.
- “The federal government has quietly launched a $67.6-million fund for digital publications covering ‘core news’ content,” reports Mumbrella. No idea how it’ll work yet.
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.
Elsewhere
- “When Google and Amazon negotiated a major $1.2 billion cloud-computing deal in 2021, their customer — the Israeli government — had an unusual demand: agree to use a secret code as part of an arrangement that would become known as the ‘winking mechanism’,” reports the Guardian. The aim? To sidestep legal obligations in other countries.
- “As tech companies build data centres worldwide to advance artificial intelligence, vulnerable communities have been hit by blackouts and water shortages,” reports the New York Times (gift link).
- In a recent video essay, the inimitable Patrick Boyle asks Is AI’s circular financing inflating a bubble?. I mean, yes?
- And when this bubble bursts, it’s going to be big. Chipmaker Nvidia is now “worth” US$5 trillion, and the New York Times outlines what that means (gift link).
- OpenAI, best known for creating ChatGPT, is now a for-profit company. At Pivot to AI, David Gerard discusses what happens next.
- The largest study of its kind shows that AI assistants misrepresent news content 45% of the time regardless of language or territory, the BBC reports. The Beeb led this research in conjunction with the European Broadcasting Union.
- A survey in the UK has found that teenage boys are using “personalised” AI for therapy and romance.
- Chatbot site Character.ai is cutting off teenagers from having conversations with its virtual characters. As we mentioned last week, Australia’s eSafety Commission had already asked them what they’re doing to protect the kids.
- China’s Ministry of State Security has deployed an AI anchor in full uniform to deliver national security warnings on social media. As the China Media Project writes, “With its new AI makeover, informing on your co-worker in the next cubicle never looked so congenial”.
- Former Intel CEO Patrick Gelsinger, now executive chairman of Gloo, wants to build a Christian AI to “hasten the coming of Christ’s return”.
- With web crawlers now wanting to ingest content for AI training as well as web search, and publishers wanting that not to happen, we have an arms race.
- 1X Technologies has launched Neo, which it says is “the world’s first consumer-ready humanoid robot designed to transform life at home.” As you’ll see in the annoying video, it’s delivered in a pod. But as this Wall Street Journal video shows, currently it’s all done by a human. Yes, “expert mode” is in fact a remote worker, somewhere, operating the robot so it can learn things. I’m sure they’ll be paid the same as a cleaner who’s physically present.
- Elon Musk is also building humanoid robots, but as Oligarch Watch asks, are they really worth a trillion dollars?.
- Amazon has released a technical summary of what happened in that massive AWS outage we discussed last week. As friend of the cybers Justin Warren wrote in his excellent free newsletter The Crux (registration required), “To its [Amazon’s] credit, the post-incident summary is quite detailed and helps people to understand how this kind of failure can happen... This is how complex systems fail. More companies could follow AWS’ example and release this sort of detail when they have failures”.
- Former employee of the Australian Signals Directorate (ASD), Peter Williams, has pleaded guilty to selling cybersecurity exploits to Russia.
WHY NOT LISTEN TO OUR PODCAST ABOUT MUSIC? My good friend Snarky Platypus and I produce a music podcast, Another Untitled Music Podcast. Look for it in your podcast app — because a new episode will be posted this Sunday.
Inquiries of note
- Treasury has posted draft regulations for the payment systems modernisation. Submissions close 11 November.
- There’s a new Senate inquiry into Triple Zero service outages. Submissions close 25 November, but before then there’s a public hearing this Monday 3 November. This is separate from the new legislation passed this week, the * Telecommunications Legislation Amendment (Triple Zero Custodian and Emergency Calling Powers) Bill 2025*.
What’s next?
Parliament continues this coming week with both houses sitting Monday to Thursday.
As usual, the government may well introduce new legislation at any time, depending on the demands of the news cycle and, to a lesser extent, the effective government of the Commonwealth.
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.