30 October 2026

Welcome

Earlier this week I visited my GP for a routine medical consultation. “May I have your permission to use an AI scribe?” he asked.

“Certain not,” I replied.

Not because I’m against transcription software per se, because I use it myself, but because in that moment I didn’t feel like I could give informed consent.

Where is this data being stored? How is it being secured? Who has access to it? What precautions are there that mistakes are detected and corrected? Would the summary put into my medical notes contain only the relevant medical information, not any more personal information that we might discuss along the way?

As it happens, I’ve been seeing the same GP for more than 25 years. Their practice is technologically literate and takes privacy seriously. I’m reasonably confident they’ve thought it through.

Nevertheless, the decision to use AI in such a sensitive environment should be a joint decision. And there needs to be more flexibility than “We’re using AI now. Agree to it or look elsewhere.”

On the other hand, people are increasingly using chatbots for their health. At least a GP’s AI scribe is likely to be more attuned to the health context and do the right thing. Probably. Maybe. Can we ever say the same about free public chatbots?

Trying to care about the Anthropic IPO, failing

Anthropic, maker of the Claude series of AI models, filed the relevant documents for its IPO, but did it really? Pivot to AI went through the 335-page S-1 form and saw that all the key numbers are missing.

I’m finding it hard to get excited by what is clearly bullshit. Still, if you want more there’s further thoughts at The Conversation, and everywhere across the internet.

Meanwhile, Google’s third-quarter earnings report came out. Their AI Overview has damaged the ad revenue from search results, and Alphabet’s share price crashed 12%.

PLEASE CONSIDER SUPPORTING THIS NEWSLETTER: The Weekly Cybers is currently unfunded. It and my podcasts The 9pm Edict and Another Untitled Music Podcast are supported by their audiences. Right now, this season’s crowdfunder is hoping to reach its first target by Sunday. It’d be lovely if you threw in your support. If you’ve already done so, thank you very much.

Also in the news this week

• “Almost three-quarters of public servants are using AI in their jobs. But concerns remain about whether it’s being used safely and effectively,” reports The Mandarin. “Commonwealth public servants were slightly more likely to be users, at 74% compared with 66% of state respondents.”

• The government’s algorithm for determining who gets what aged care support lacks safeguards and has led to unfair outcomes for elderly patients, according to a report (PDF) by the Commonwealth Ombudsman. While the government has accepted six of the ten recommendations, the one about having some sort of human override was merely “noted”. Which of course is code for “will be ignored”.

• On a related note, in a conversation earlier today I was reminded that the EU’s Artificial Intelligence Act includes a Right to Explanation of Individual Decision-Making. No black-box decisions!

• The Digital Duty of Care Bill was passed in the House of Representatives with a number of amendments. One was to ban certain chatbot activity, including “material or conduct that simulates emotional intimacy and encourages or fosters a child’s dependence on it in place of human relationships”. The bill now goes to the Senate, where its fate remains uncertain as both the Coalition and One Nation oppose the bill. Apparently it’s some sort of “Orwellian” system of “censorship”. The committee inquiry into the bill reports back on 20 November, leaving just four sitting days to get the thing passed before Christmas. I’m certain this will lead to thoughtful legislation, just as it did with the social media age restrictions.

• Following the failure of the Firmus datacentre IPO earlier this month, the Australian Securities ​and Investments Commission (ASIC) has some questions for the banks behind the planned float.

• AI is killing off entry-level cybersecurity jobs, reports Information Age. Meanwhile, one in five borrowers have used AI to fake mortgage documents.

• From last Friday, news that the ABC has wiped some staff devices that were potentially exposed to a spyware campaign from an Israeli company.

LATEST PODCAST: With so much happening in the digital realm, I thought it’d worth getting some expert commentary. So last week I recorded The 9pm Super Intelligent Duty of Care with Johanna Weaver, executive director and co-founder of the Tech Policy Design Institute in Canberra. In this episode we discuss the Digital Duty of Care Bill, why it isn’t an “AI race” between the US and China, the law relating to rogue AI (there’s no such thing as “rogue AI”), and of course whether artificial intelligence is really “super intelligence” (it isn’t). All that and much more. Look for The 9pm Edict in your podcast app.

Elsewhere

• In breaking news, cryptocurrency exchange Binance has frozen funds after a potential cyber attack. As I write these words, around US$3 billion of assets have been frozen, and the price of Bitcoin has dropped about 5%. We don’t yet know whether there have been any confirmed losses. Watch this space.

• A cyber attack against Moldova took more than 80 government services offline. “The hackers stole extremely sensitive details on 750,000 citizens, which is roughly a third of the country’s total population,” reports Risky Business News.

• Elon Musk recently slowed down Ukraine’s access to Starshield, the military communications service running on its Starlink satellites. The Pentagon inspector general’s office has declined to investigate, despite a request from US senators. There’s some interesting questions here about the relationship between government and the private sector suppliers when it comes to military and foreign policy decisions.

• “Despite massive investments and widespread hype around artificial intelligence, the technology has delivered zero financial payoff to corporate America outside of tech,” writes Torsten Slok, chief economist at Apollo Global Management. “AI does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness, all of which shape reputational risk and ultimately drive revenue growth and profit margins.” There are charts.

• OpenAI’s ChatGPT promised a suicidal man that human help was coming, but it wasn’t. The man did not survive.

• In San Francisco, you can earn US$20 an hour taking robot dogs for a walk to teach them how to get around the city. No experience needed, just the ability to carry the 15-kilogram robot back to base if it overheats and carks it.

• Via Boing Boing, “AI-written Disney guidebooks send tourists to rides that don't exist.”

• And finally, well, I’ll just copy-paste a couple of sentences. “Meet the Spank Vault: a rugged, solar-powered, offline porn vault that ships pre-loaded with thousands of genres and niches going back as far as the 1970s, and makes endless new AI mash-ups with no internet... One rugged hard case holds everything you need to keep the porn flowing after the internet dies.”

OPENAUSTRALIA FOUNDATION NEEDS YOUR SUPPORT: This free public online library of what Australian governments do is running out of money and is in danger of closing. If you use Planning Alerts, They Vote for You, Right to Know, or OpenAustralia.org.au, or just think they deserve to exist, please support them, especially if you’re an organisation that uses their services.

Inquiries of note

• Treasury has two inquiries running in relation to the Consumer Data Right, the right to share your data between services providers that currently allows you to more easily switch banking and energy providers. One is about draft standards for minimum baseline authentication the other is draft standards for the Financial-Grade API (FAPI) 2.0 security profile. Submissions close 26 November.

What’s next?

Parliament is now on break for a bit. The Senate returns on Monday 12 November, and the House of Reps joins them on Monday 23 November. Parliament is scheduled to wrap up on Thursday 26 November, but of course this may change.

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.