23 January 2026

Welcome

It’s a smaller edition this week, and it’s mostly links to stories elsewhere. Consider it a treat before the long weekend.

Amazingly, there’s nothing about the teen social media ban, at least not ours in Australia. The UK is having a go, however.

Yes, the hate speech laws were passed

You’re probably sick of this news already, so let’s get it out of the way. The government split the post-Bondi knee-jerk legislation into two halves, both of which were passed by parliament on Tuesday.

These were the Combatting Antisemitism, Hate and Extremism (Firearms and Customs Laws) Bill 2026 and the more controversial Combatting Antisemitism, Hate and Extremism (Criminal and Migration Laws) Bill 2026.

Weirdly, the latter bill led to the Coalition’s collapse. While amusing for some and horrifying for others, that’s outside the scope of this newsletter.

How all this will work in practice remains to be seen. However there are warnings that there may be unintended consequences, such as legitimate political commentary being classed as hate speech.

The (government) devil is in the details

While the focus of parliament was on those two pieces of legislation, it’s worth remembering that well over 300 documents were tabled. Few journalists take the time to go through them, although that’s generally not their fault. There’s only so much time in the day.

As always, many of these documents were routine. Things like annual fisheries quotas, rules for determining how specific medical conditions might be blamed on various causative factors, and the annual management plan for Mawson’s Huts.

However we also learned more about how the telecommunications outage register will work, more about the MyHealthRecord share by default rules for service providers, that Corrective Services NSW is now an enforcement agency under the Telecommunications (Interception and Access) Act, and new rules for how ASIO can demand lawyers keep information secret.

Your humble writer suggests that more attention might be given to the actual mechanics of government rather than the boasts and insults of the politicians. They’re all listed here.

CAN GROK REALLY “APOLOGISE” FOR ANYTHING? In my latest podcast I chat with a philosopher who’s deeply interested in these questions and many others, such as chatbots that mimic the dead, and technological determinism: The 9pm Unnatural Act with a Pig with Patrick Stokes. Don’t let the title distract you. Look for The 9pm Edict in your podcast app.

Also in the news

• The Australian Taxation Office (ATO) contact centre is getting into AI as part of “a massive, five-year contact centre system transformation project, codenamed RISE – short for Reimagined Interactions Streamlined and Effective”.

• A Tasmanian tourism company used AI to generate content for its website. Hilarity ensued. “I’ve seen it create animals I’ve never seen before — three-legged wombat, crocodile-looking things,” and non-existent tourist attractions.

• The Information Commissioner wants more transparency regarding automated decision making in government. Robodebt was probably just the top of the iceberg.

• The Privacy Act gives you right right to access the personal data that organisations hold about you, but getting that data is difficult.

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

• Microsoft boss Satya Nadella really wants us to use AI. He’s worried about it just being a bubble. Gosh.

• From the Lowy Institute, analysis of the Grok bans and how Indonesia, Malaysia, and the Philippines are setting the pace on global AI governance norms.

• From New Zealand, “I spent 18 months training generative AI – here’s what I learned.”

• Professor of plant sciences Marcel Bucher turned off ChatGPT’s “data consent” option and two years of academic work vanished. It sounds like extremely poor interface design, but he did ask OpenAI to stop storing his data, so they stopped storing his data.

• Iran shut down the internet to help suppress domestic unrest, but activists and civil society groups had already smuggled in 50,000 Starlink terminals.

• “The Gates Foundation and OpenAI are setting up a $50 million partnership to help several African countries use artificial intelligence to improve their health systems,” reports Reuters.

• The UK government is starting a consultation on whether to have its own ban on social media for under-16s. Indeed, the House of Lords has said it has to happen.

• A magnificent headline but a sad story: “A man bought Meta’s AI glasses, and ended up wandering the desert searching for aliens to abduct him.”

• In a sign of the times, WIRED seems to have removed the paywall from this one: How to Protest Safely in the Age of Surveillance.

• A fascinating list of Falsehoods People Believe About Computers. The first one establishes the vibe: “The current technology systems exist mostly to serve my needs and are not deliberately and consciously optimised to hurt me.” umber 8 could well apply to robodebt: “All technology harms are mitigated once someone has been blamed, and there is nothing useful to be learned after that.”

Inquiries of note

There’s a couple relevant inquiries which slipped past me.

• The House of Representatives Standing Committee on Economics has an inquiry into current digital payment schemes and emerging technologies. Submissions close next Friday 30 January.

• There’s also an inquiry into the Standing Orders relating to petitions, which might seem boring but it’s about “the impact of emerging technologies, including artificial intelligence and automated systems, on the integrity and authenticity of the petitions process”, among other things. Submissions close 6 February.

What’s next?

Parliament is now on break for a week, returning for three days on Tuesday 3 February. The following week the House of Reps will sit for four days while there’s some supplementary Senate Estimates hearings.

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.