27 March 2026

Welcome

It’s a week when too much news is barely enough, to use a well-known Australian expression, and once more I’m having to leave most of the lengthy writing for others.

Meta and YouTube having to pay damages resulting from their “addictive” social media sets a vitally important precedent. I can’t help but feeling the tide is turning against the until-now invincible big-tech platforms.

Among the other stories this week, I did enjoy the one about AI trying to tell the difference between a crocodile and a log. After all, even humans have trouble with that one.

Social media is addictive, says US jury

Yes indeed, Meta and YouTube have designed addictive products that harm young people, according to the jury in a landmark trial in Los Angeles.

After nine days of deliberations, the tech companies were found to have been negligent, and to have failed to provide adequate warnings about the potential dangers.

The 20-year-old plaintiff, referred to as KGM, had claimed she became addicted to YouTube at age six and Instagram at nine, leading to a number of subsequent harms including anxiety, depression, body dysmorphia, and suicidal ideation.

KGM was awarded US$6 million in damages, with Meta to pay 70% and YouTube 30%.

Both TikTok and Snap, the latter being the maker of Snapchat, settled on confidential terms to avoid being included in this trial.

“KGM’s case is the first of its kind, but won’t be the last: it is one of more than 20 ‘bellwether’ trials due to go to court soon. These are essentially test cases used to gauge juries’ reactions and set a legal precedent,” writes Rob Nicholls, senior research associate in media and communications at the University of Sydney.

“As such, the verdict is set to have far reaching ripple effects. It could be big tech’s big tobacco moment, with thousands more similar cases waiting in the wings.”

There’s plenty more expert reaction at Scimex.

Meanwhile in New Mexico, Meta was ordered to pay US$375 million in civil penalties after being found liable in child exploitation case. Once more, the company was found to have misled consumers about the safety of its platforms.

Australian law firms are considering their options.

Harmful social media tools cop Australian age restrictions

Perhaps in response to the US findings, communications minister Anika Wells introduced new rules which will age-restrict specific functions of social media platforms.

According to the minister’s press release, platforms will be defined as age-restricted if they have a so-called “recommender systems”, or algorithms “designed to be addictive and provide constant dopamine hits by providing highly personalised material based on information associated with a young person’s account”.

This includes any of these three features is a user is logged in:

• endless-feeds (infinite scroll), designed to keep young people engaged for as long as possible by showing new content as you scroll, swipe or flick with no end point

• feedback features (such as displaying the number of “likes” or “upvotes” a user has received), designed to give young people quick, low-effort feedback and create pressure to seek approval and compare themselves to others

• time-limited features (such as disappearing “stories”) designed to create urgency so young people check apps constantly out of fear of missing out.

Again, they’re the minister’s words. The actual legislative instrument is worded more briefly.

“The new rule has no impact on platforms eSafety has previously assessed as out of scope (on the basis that they are excluded under the legislative rule made in July 2025), including Discord, Google Classroom, WhatsApp and Roblox,” the press release says.

Also in the news

• The government has issued a new set of Expectations of data centres and AI infrastructure developers, who are now expected to “consider and contribute to Australia’s interest, which includes Australia’s national security and data sovereignty”, among other things. There’s some analysis at The Conversation.

• Australia and the EU have signed a cyber and maritime defence partnership. The Australian Strategic Policy Institute (ASPI) says Europe’s defence push needs to be taken seriously, and has published an explainer.

• The government has introduced the Australian Criminal Intelligence Commission Bill 2026, which will “establish the ACIC as Australia’s national criminal intelligence agency, realigning its functions and powers so that the ACIC can target its efforts towards producing unique, actionable and insightful intelligence relevant to serious and organised crime”. Its 293 pages represent what looks like a complete legislative overhaul of the organisation, though of course I haven’t had time to read it.

• They also introduced the Crimes and Other Legislation Amendment (Omnibus No. 1) Bill 2026, which includes all manner of things, from how drug quantity thresholds are calculated, to classifying the new Western Sydney International Airport as a major airport for various law enforcement purposes. There’s also some minor changes to various surveillance legislation.

• The Independent Review of the Security of Critical Infrastructure Act 2018 found that the act “requires major legislative change to remove complexity and confusion while becoming more agile and responsive”. One key recommendation is “[moving] from a ‘light touch’ compliance approach with a focus on administration and documentation to that of a penalty-based risk management process with the real enforcement of penalties”. In other words, the laws are toothless.

• Operators of low earth orbit (LEO) satellites could deliver internet connectivity without having an Australian presence, which could put them outside Australian data laws.

• The Australian Communications and Media Authority (ACMA) is going to design its own enforceable industry standard to replace the industry-developed Telecommunications Consumer Protections Code which no one but the telco industry liked, although they don’t seem to have formally started the consultation process yet.

• A 78-year-old Melbourne man has been sentenced to six years in jail after pleading guilty to charges of grooming children online and hosting a child sexual abuse material site.

• The ACCC has hit the owner of PhotobookShop with $39,600 in penalties for alleged misleading influencer reviews on Instagram. An influencer blew the whistle when they saw a contract clause, “Please ensure that your videos do not mention that the product is free, sponsored, or that PhotobookShop contacted you to create them in exchange for products.” Sharyn Smith, founder and CEO of influencer agency Social Soup, reckons a $40k fine is a rounding error, given the “untold damage” to the influencer market.

• “The company responsible for the highly criticised Bureau of Meteorology website overhaul has won a $16 million tender to develop another climate data website for the government,” reports ABC News, and climate scientists are worried.

• Australia is flooded with climate misinformation.

• A project in Queensland is trying to detect crocodiles using AI but, just as for humans, the problem will be telling the difference between a croc and a log.

• OpenAI will soon start running adverts in ChatGPT in Australia.

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

• In the UK, the House of Lords has once again called for a teen social media ban similar to Australia’s. The proposal would give ministers 12 months to decide which social media platforms should be included.

• The EU is formally investigating Snapchat over child protection concerns.

• The US has banned new foreign-made consumer internet routers over national security concerns, although Americans can continue using the routers they have.

• Wikipedia has banned the use of AI for writing on the site.

• Ars Technica has kicked off a three-part series on this idea of orbital data centres by asking There’s no way this is economically viable, right?.

• OpenAI has shut down Sora, its AI video-generating app, and cancelled its billion-dollar partnership with Disney. OpenAI is reported going to dump its focus on video altogether.

• An interesting piece at The Mandarin: Policy by trial and error: How Silicon Valley culture has infiltrated governments.

HOW’S THAT AI BUBBLE GOING? My most recent podcast episode is The 9pm S-Bend of Technology with David Gerard. He’s the editor of the Pivot to AI newsletter, video essay, and podcast. Look for The 9pm Edict in your podcast app

And I know that’s been the latest episode for a few weeks now, but I’ll start clearing the backlog of already-recorded episodes very soon.

Inquiries of note

• The government is reviewing the National Classification Scheme, the system under which content is classified or “rated” under the familiar labels of G, PG, M, and so on. The system covers films, video games, and certain publications. Submissions close 4 May at 1155pm AEST.

What’s next?

Parliament returns for three days of sittings from this Monday 30 March, the short week before Easter.

The House of Reps program shows planned debate on the Telecommunications Amendment (Enhancing Consumer Safeguards) Bill and the Telecommunications Legislation Amendment (Universal Outdoor Mobile Obligation) Bill.

At this stage the Senate program shows nothing of interest to this newsletter.

As always, the government may introduce more legislation as suits their needs on the day.

Finally, because it’s Easter next weekend, this newsletter will appear on Thursday 2 April rather than Good Friday. See you then.

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.