Australia is looking at making social platforms let people switch the algorithm off — a much bigger swing than the teen account ban the country spent most of the last year building.
Social services minister Tanya Plibersek told the ABC's Insiders that the government is weighing several approaches to algorithmic amplification, opt-in and opt-out both on the table. The Guardian reported it on Sunday.
What set this off
The context is bleak. Plibersek pointed to a recent teenage strangling death and rising reports of sexual violence among teenagers when she explained why the government is now looking past account bans at the recommendation systems themselves.
"If you set up a social media account as a young boy in Australia today, it takes an average of 23 minutes for you to be fed misogynistic, violent misogynistic content online," she said — and called that a deliberate choice by the companies, not an accident.
The change of target is the real story. The under-16 rules were built to keep kids off the platforms full stop. So far there's little sign they've shifted the behaviours they were meant to fix. Going after what the feed does, instead of who's allowed to open it, is a different fight.
The evidence is not thin
The case against engagement-ranked feeds has been piling up for over a decade. Harvard Business Review found in 2016 that high-arousal emotions — anger and fear especially — drove the biggest response and the most sharing. A 2023 paper in the National Library of Medicine found negative words in headlines pushed consumption up.
A 2021 Science Direct study of Instagram found posts that triggered negative emotions pulled up to 70% more likes than positive ones. KAIST researchers, published in 2023, found angry posts travelled further than anxious ones and angry users were likelier to hit share. And back in 2012, Wharton found the angrier a post made people, the further it spread.
You don't need to assume bad intent anywhere in that chain. Ranking systems count likes and comments and don't ask what caused them. Content that makes people furious gets engagement, engagement buys reach, and round it goes. Add Pew's finding that most people now get their news through these apps and the stakes stop being abstract.
Why Meta will dig in
Meta has spent years arguing that algorithmic matching is what keeps people engaged — which is, conveniently, what makes its ad inventory worth anything. A law letting users turn recommendations off goes straight at that. Expect the fight to get reframed as a trade issue rather than a child safety one.
There's already a half-precedent. In the EU, Meta lets you pay a monthly fee to get out of algorithmic matching. Hardly anyone does, which tells us almost nothing about how a non-algorithmic feed performs at scale.
The version worth running
An opt-out tucked three menus deep will land exactly where the EU version landed: a handful of people who go looking, and no change at all for everyone else. Defaults decide these things.
The genuinely interesting idea is the one Plibersek mentioned almost in passing — make it opt-in. Recommendations off unless you switch them on, leaving a chronological feed of accounts you actually chose. That would be the first proper read on what happens to engagement, to hours spent, and to what teenagers get served when the recommendation engine isn't running by default.
Australia's small enough that the platforms could swallow the experiment, and it's already shown it'll legislate before bigger jurisdictions do. Nothing's drafted, and the government is still calling this exploratory. But the country that turned teen bans into a global talking point is asking a harder question now, and plenty of regulators will be watching what comes back.
Image: Kaboompics, via Pexels
