Australia wants to hand social media users something the platforms have rarely offered without strings attached: a simple way to stop the algorithm from choosing what they see. A package of tech laws introduced to the federal parliament this week includes a so-called algorithm switch, which would oblige social media companies to let people turn off recommended content and see only posts from the accounts they actually follow.
It sounds like a small design change. In practice it goes to the heart of how modern platforms make money. Recommendation engines decide which videos, posts and clips fill a feed, and they are built to keep people scrolling for as long as possible. Critics have argued for years that this logic favours whatever triggers the strongest emotional reaction, whether that is outrage, fear, envy or insecurity, and that vulnerable users pay the price.
Prime Minister Anthony Albanese has framed the plan as a matter of personal choice rather than state control. His message was essentially that anyone who wants to doom scroll is free to do so, but that people should be able to decide for themselves. The government stopped short of forcing the switch to be off by default, which is what campaigners had originally asked for.
The switch is only one part of a much broader Digital Duty of Care bill. Under that law, online services of almost every kind, from social networks and gaming sites to chatbots, would have to take reasonably practicable steps to provide a safe environment. Children would have to be shielded from harmful material and design features, including pornography and content that promotes eating disorders or hostile attitudes towards women. Adults would be protected from a narrower list of seriously harmful and illegal material. Companies that break the rules could face fines of up to 100 million Australian dollars, roughly 70 million US dollars.
Communications Minister Anika Wells has argued that the duty of care fills many of the gaps left by Australia’s earlier law setting a minimum age for social media. Instead of focusing only on who is allowed on a platform, it puts responsibility on the companies that profit from Australian users, rather than on the people who get hurt.
Real stories behind the policy, and the doubts that remain
The push did not come out of nowhere. Personal accounts of being dragged into harmful content have shaped the debate. One young woman, now in her mid-twenties and recovering from anorexia, has described how her feeds as a teenager filled up with extreme workout videos, dangerous calorie challenges and strangers documenting what little they ate. Even today, with every safety setting enabled and a long list of blocked words, she says weight loss content still finds its way to her.
A young man from Melbourne has told a similar story from a different corner of the internet. At a low point in his life, before he was diagnosed with bipolar disorder, his feeds were flooded with so-called manosphere influencers. He went viral with angry videos of his own and only found his way out after a mental health crisis and a stay in hospital. Both say social media was not the only cause of their worst years, but that it played a major role.
Their experiences fit into a growing body of research. Studies have linked heavy social media use to brain responses similar to those seen in gambling, found that algorithms push unsolicited content that can damage mental health, and connected recommendation systems to the spread of disinformation and polarisation. Lawsuits and whistleblowers have added claims that companies understood the risks and accepted them because engagement drives revenue. The platforms counter that the dangers are exaggerated and that they already offer effective safety tools.
The campaign that put the switch on the political agenda was called Fix Our Feeds. It was launched by Chanel Contos, an activist already known for helping to reform consent education in Australian schools. Her idea was a return to the social media of the early 2010s, when feeds were mostly made up of friends and accounts people had chosen, and when they actually came to an end. More than 15,000 people signed her petition, and she won support from two former prime ministers, several members of parliament, mental health advocates and online safety experts. Contos now hopes to take the campaign to Canada and the United Kingdom, arguing that if several like-minded countries move together, the change could reach a critical mass.
Big tech has so far reacted cautiously. Meta, Snap, TikTok, X and Google declined to comment publicly on the legislation. Instagram chief Adam Mosseri, speaking in Sydney, argued that non-algorithmic feeds would actually show users fewer posts from their friends and rejected the idea that his app is addictive, drawing a distinction between casual binge use and clinical addiction. Right-wing parties in Australia and the Trump administration have warned that the bill could lead to over-moderation and censorship. Wells has responded that adults would remain free to follow whoever they like and post whatever views they hold.
The bigger question may be whether the rules can be enforced at all. Australia is still struggling to implement its ban on social media for children, with surveys suggesting most young users who were on the platforms before the ban are still there and no fines issued so far. The European Union’s experience is also sobering. After the EU introduced algorithm controls in 2023, platforms did add following-only feeds, but often in a clumsy form. On Instagram, for example, switching off the algorithm has meant losing easy access to messages, the profile page and settings, and the feed snaps back to the algorithmic version the next time the app is opened. Only in the Netherlands, after a court ruling against Meta under the Digital Services Act, do users keep a lasting choice.
Canberra says it has learned from this. Its draft rules state that following-only feeds must not be harder to use or less functional, that the setting must be obvious, and that it must not change without an explicit decision by the user, even after leaving the app.
Public opinion appears split along generational lines. Polling earlier this month suggested that about two thirds of Gen Z users would keep their algorithms switched on, while nearly half of Australians overall said they would turn them off. Even some of the people whose stories inspired the reform are unsure how much it will change. But for those who know how quickly a bad day can turn into a spiral, the simple option to step out of the recommendation stream may be the point. Whether the platforms make that option genuinely easy to use will decide if Australia’s experiment becomes a model for others or another setting most people never touch.
