The Weekly Cybers #58
Proposed new rules for social media platforms, calls for sweeping changes to government AI, and no, Facebook isn’t for emergencies.
Welcome
The Australian government’s plans to regulate the big online platforms continue to solidify as the companies put forward their draft codes of practice, including rules for complying with the social media age restrictions.
A parliamentary inquiry calls for sweeping changes to the government’s governance of its own AI use.
And I’m guaranteed to make a few more enemies with my lengthy comments about the use of Facebook during an emergency.
Social media platforms write their draft rules
The social media platforms have submitted their draft codes to protect Australian children from access or exposure to online pornography and other “class 2” material, which includes self-harm, eating disorder content, promotion of suicide, and violence.
As the Guardian reports, they have six months to implement the rules, assuming the eSafety Commissioner smiles upon them. They’ll then then face fines of up to $50 million if kids are found on adult sites.
In his analysis of the codes, Professor Toby Murray from the University of Melbourne says that we still don’t know which age assurance technologies will be deemed “appropriate” under the Online Safety Act.
Publicly available data suggests that even the best systems that use face pics for age estimation, for example, have an average error of 3.7 years.
“There is a real risk many teens will remain able to access online porn and harmful deepfake apps despite these new codes,” he wrote.
Meta, Snapchat, TikTok issue coordinated complaints
Meanwhile, three major players have launched a campaign against what they’ve labelled an “irrational” and “shortsighted” decision to exempt YouTube from the social media age restrictions.
Needless to say, Meta, Snapchat, and TikTok all want YouTube to suffer the same fate as them, despite communication minister Michelle Rowland’s frequent assertion that they’d be exempt.
Remember, however, the legislation itself does not list who’s in and who’s out. Everything is down to the feelpinion of the minister of the day.
Calls for sweeping changes to government AI use
A parliamentary committee has called for an AI use census down to the individual public servant level, as well as specific legislation governing AI use, mandatory rules and governance requirements, and standardised definitions and training across the public service.
The report from the Joint Committee of Public Accounts and Audit, titled Proceed with Caution, expresses “very grave concern” that if the government doesn’t get things under control now, “this technology will outpace its ability to do so in the very near future”.
It also recommends establishing a statutory Joint Committee on Artificial Intelligence and Emerging Technologies.
IF YOU’VE FOUND 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.
No, Facebook isn’t for emergency communication
Meta copped some flak this week when Facebook blocked search results for “Cyclone Alfred”, and YouTube reportedly sometimes demanded a login before showing updates from the Bureau of Meteorology (BoM).
This embarrassing behaviour led to comments such as those of Cam Wilson at The Sizzle ($):
“I know it’s easy to default to private, seemingly ubiquitous, platforms that might even be better than what governments offer. But governments need to invest in creating and running the platforms — preferably open platforms! — because they can’t leave serving the public to the goodwill of private companies. You shouldn’t build your emergency broadcast systems in other people’s kingdoms.”
I agree with all of that except for the last sentence. Or rather, I also agree with the last sentence, but is that really what happened here?
Not all comms during an emergency is “emergency communication”
Let me clarify what I’m talking about. Searching for posts on Facebook isn’t “emergency communication”. Neither are the BoM’s general weather forecasts.
Emergency communications are those urgent messages aimed at people who are directly affected. We’re familiar with these messages in relation to bushfires. Watch and act. Evacuate now. Such and such highway is closed.
We already have government-owned and run emergency broadcast services for this: the ABC, which is the nominated emergency broadcaster, and Emergency Alert, which sends geotargeted SMS messages and calls from +61 444 444 444.
There’s also the official apps such as NSW’s Hazards Near Me, which provide geotargeted alerts.
If someone’s phone is still working, then they still have access to emergency information. And of course the emergency workers have their own dedicated radio channels.
At a lower priority, though, is what I call “communications during an emergency”. These are all the other communications to and from the emergency-stricken areas. An obvious example would people from outside the area checking on friends or relatives, or people leaving the area calling ahead to make arrangements.
It might seem cruel to say this, but these are not emergency messages. It’s certainly a worry if you can’t contact your friends or relatives in a flood zone, but that’s about your anxiety, not their safety. And in any event, you don’t send and receive these messages by searching on Facebook.
Finally there’s what we might call “news about the emergency”, for consumption by people who aren’t really affected by the emergency itself but just want to know what’s going on. Well, that’s what the news is for, and if you don’t find out for a few hours or more, so what?
So the question in my mind is this: What do we expect each platform to be able to do during an emergency?
My view is that it isn’t and shouldn’t be a social media platform’s job to be an emergency communications channel. We already have these channels, and they’re already available on everyone’s phone.
Facebook is no more an emergency broadcast system than an ice cream van is a fire truck. Stop wanting every platform to be everything for everybody.
This is not to excuse Meta’s problems, of course
Still, Facebook’s search function fell over during a period of extreme interest in “Cyclone Alfred”. Oops.
Now it may be that a sudden increase in searches for a specific thing looked a lot like “inauthentic behaviour”, perhaps like bots trying to artificially inflate the interest in a particular post and game the trending topics.
In my view Meta should have been able to figure out what was going on well before the media started asking questions. Perhaps this points to a lack of human judgement in Meta’s processes.
We still don’t know exactly what happened beyond “technical issue”, and I doubt we ever will given Meta’s lack of transparency.
And of course misinformation and disinformation is a problem. Given its dominance of people’s online mindsets, Meta still needs to lift its game.
Also in the news
- The Commonwealth Ombudsman released its annual Oversight of Covert Electronic Surveillance report for 2023–2014. In the media release (PDF) it says: “Two common issues were insufficient record keeping and a lack of general training and guidance material for officers. Not only were these common issues, but for some agencies they are also repeat findings across several inspection periods”.
- The Greens proposed a 3% revenue tax on big tech, which modelling suggests would raise $11.5 billion over the next decade.
- According to the Australian Medical Association (AMA), one-in-four GPs are using AI to assist with note-taking. Currently Australia has no mandatory guidelines for the use of AI in medical settings.
- AI doesn’t really “learn”, and knowing why will help you use it more responsibly.
- From the Nine papers, “Australian publisher Black Inc. is facing widespread criticism after asking its authors to consent to their works being used to train artificial intelligence”.
- On a related note, UNSW’s Professor Toby Walsh writes: “I have been an AI researcher for 40 years. What tech giants are doing to book publishing is akin to theft.”
- Something I’d missed last week: the Office of the Australian Information Commissioner (OAIC) has released a Digital ID regulatory strategy which sets out how the agency proposes to “build trust and confidence in Australia’s Digital ID System and make identity verification in Australia more secure and privacy protective”. I still have to get around to reading it.
- The defence department has confirmed it’s exploring the possibility of using SpaceX’s Starshield military-grade communications satellite system, reports InnovationAus ($).
- Audio artists working on video games are not happy with AI taking their jobs.
- A Melbourne start-up has launched a “biological computer” made of human brain cells.
- Telstra copped a $626,000 penalty and offered a comprehensive court-enforceable undertaking for sending texts that didn’t comply with anti-spam laws.
Elsewhere
- Apple is taking legal action in an attempt to overturn the UK’s demand for backdoor access to users’ encrypted data. The new US director of national intelligence, Tulsi Gabbard, says the UK’s demand is a “clear and egregious violation of Americans’ privacy and civil liberties”.
- Microsoft is shutting down Skype on 5 May. You’ll have to option of moving everything to Teams. Or, you know, not.
- From 404 Media, “Buying a $250 residency card from a tropical island let me bypass US crypto laws”. The island in question is the nation of Palau, which is actually an archipelago of over 500 islands.
Inquiries of note
Nothing new for us this week.
What’s next?
Parliament is currently on a break. Both houses return for three days of sittings on Tuesday 25 March, then the Senate only returns for two weeks starting Monday 7 April. Unless the election is called. I explained the potential timing for that three weeks ago.
Update 8 March 2025: The current theory is that a federal Budget will be delivered on 25 March, and the election will be held on one of the first three Saturdays in May.
DOES SOMETHING IN THE EMAIL LOOK WRONG? 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 look at 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 specifically a cyber *security* newsletter. For that that I recommend Risky Biz News and Cyber Daily, among others.