For decades, unregulated recommendation algorithms on social media have faced growing backlash over their role in amplifying harmful content that damages user mental health, fuels extremism, and erodes public discourse. Now, Australia is pushing forward with ambitious new legislation that could rewrite the rules of global social media design, giving users the explicit power to opt out of algorithmic feeds entirely and only view content from accounts they choose to follow.
The story of this reform begins with the personal stories of harm that pushed lawmakers to act. Varsha Yajman, now 24 and recovering from anorexia, first encountered dangerous eating disorder content as a 13-year-old new to social media. Her feed became flooded with endless workout clips promising unachievable body standards, strangers promoting lethal calorie challenges, and public profiles that turned shared hospitalizations for eating disorders into a toxic competition. Every night before bed, she would scroll through daily food diaries posted by strangers, convincing herself she did not need to eat because she could watch others do it instead. Even now, years into recovery and with every safety setting activated and hundreds of harmful keywords blocked, Yajman says weight loss content still appears on her feed on a regular basis.
Flynn Martin, a Melbourne resident who was pulled out of the misogynistic “manosphere” after a psychotic break and hospitalization, shares a similar experience. Four years ago, at the age of 20, Martin was struggling with bipolar disorder and deep life insecurity when algorithmic recommendations began pushing content from self-described misogynist influencer Andrew Tate and other extremist manosphere figures onto his feed. He quickly became trapped in an echo chamber of radicalizing rhetoric, buying into the idea that society was stacked against men and that the only path forward was to embody an aggressive “alpha male” persona. Like Yajman, Martin acknowledges social media was not the sole cause of his struggles, but says it was a critical accelerating factor that pushed him into a dangerous place he struggled to escape.
These are not isolated incidents. Research from across the globe has linked algorithmic content curation to a range of harms: studies show social media use triggers brain responses similar to gambling and drug addiction, unsolicited recommended content regularly pushes content that worsens user mental health, and algorithms amplify disinformation and political polarization that can lead to offline violence. Whistleblower testimonies and court documents have repeatedly alleged that big tech companies are fully aware of these harms, but choose to overlook them because algorithmic amplification drives higher user engagement and bigger profits. As consent education activist Chanel Contos, who led the grassroots campaign for reform, explains: “The result is that they promote content that is outrageous or that invokes strong emotions from people – so it might be anger or fear or desire or lust or insecurity, whatever it is, it’s going to keep us scrolling.”
Contos, who successfully pushed for reform of Australian sexual consent education, launched the “Fix Our Feeds” campaign in 2021 after identifying algorithmic harm as a systemic public issue. Her original proposal called for algorithmic recommendation to be turned off by default, with users required to actively opt in if they wanted personalized feeds, a return to the early 2010s model of social media where feeds only showed content from accounts users chose to follow. More than 15,000 Australians signed her petition, and the campaign earned backing from two former prime ministers, dozens of sitting MPs, mental health advocates, and leading online safety experts. In September, the Australian government announced it would adopt a modified version of Contos’ proposal as part of a broader package of online safety reforms introduced to parliament this week.
At the core of the new Digital Duty of Care bill is a mandate that would force all major social media platforms to offer users a clear “algorithm switch” that lets them disable recommended content entirely, restricting their feed to posts only from accounts they actively follow. The bill also imposes a broad legal duty of care on all major online services — including social media platforms, gaming sites, and chatbots — requiring them to maintain “as safe as reasonably practicable” online environments for users. Under the law, children must be protected from all harmful material, including content that promotes eating disorders, violence against women, and pornography, while adults must be protected from a narrower set of seriously harmful illegal content. Companies found in breach of the law face maximum fines of up to 100 million Australian dollars, roughly 70 million USD.
Communications Minister Anika Wells argues the new law shifts responsibility for online safety from harm survivors to the big tech companies that profit from Australian users, filling gaps left by earlier child social media bans. “The digital duty of care answers a lot of the ‘what abouts’ of the social media minimum age law because it’s much broader; it captures all online services and it shifts the onus of responsibility from the shoulders of victims onto the shoulders of the big tech operators who seek to make money off Australian users,” Wells told the BBC.
Prime Minister Anthony Albanese emphasized that the reform is about putting power in the hands of users, not the government. “This is not about giving government control, it is about giving people control,” Albanese said. “If people want to doom scroll… that’s fine, that’s up to them.”
The proposed law has faced criticism from multiple sides. Right-wing groups, including the Trump administration, have claimed the bill will encourage over-moderation and censorship of free speech. Supporters of the reform reject this argument, noting that users can still follow any accounts they want and share any views they choose, regardless of whether they opt into algorithmic recommendations. “If adults listen to manosphere podcasts, or people post their views about the government, or make memes, or want to say the moon landing was fake – they can continue to do so without restriction,” Wells said.
Big tech companies have largely stayed silent on the legislation, with all major firms including Meta, TikTok, X, Google, and Snap declining to comment on the bill to the BBC. Instagram head Adam Mosseri did argue against the change during a recent press briefing in Sydney, claiming that non-algorithmic feeds would reduce the number of posts from friends that users see and rejecting claims that Instagram is addictive — despite a recent U.S. jury ruling that found the platform is designed to be addictive.
Critics of the policy also point to major questions about enforcement. Australia has already struggled to implement its existing ban on social media for children under a certain age, with surveys showing most underage users who were on social media before the ban still use the platforms, and no fines have ever been issued for breaches. International experience also offers a cautionary tale: after the EU introduced algorithm control requirements in 2023, most platforms implemented the following-only feed option in a deliberately clunky, low-functionality way that discourages use. On Instagram, for example, turning off algorithmic recommendations means users lose access to direct messages and their own profile settings, and the app automatically resets to the algorithmic default every time a user closes and reopens the app. Only in the Netherlands, after a court ruled Meta violated the EU’s Digital Services Act, is a fully functional following-only feed available to users.
To avoid these pitfalls, the Australian draft legislation explicitly requires that following-only feeds cannot be less functional than algorithmic feeds, that the option to switch algorithms must be obvious and easy for users to find, and that settings cannot be changed without a user’s explicit permission, even if they leave the app.
Buoyed by the success of her campaign in Australia, Contos is now taking the Fix Our Feeds push global, targeting Canada and the United Kingdom first. If the three aligned jurisdictions all adopt similar rules, Contos says it could create enough regulatory critical mass to drive global change in how social platforms are designed.
For the people whose stories helped drive the reform, opinions on the policy are mixed. Yajman is skeptical that Australia can actually enforce the rules against powerful big tech companies, but says even having the option is a comfort. “On a bad day I’m still susceptible to all of the things that I was at 17. To know that, if I’m having a bad day, I can opt out makes me feel really good,” she said. Martin argues lawmakers need to be realistic about the limits of the reform, noting that most people use social media for entertainment, and algorithms are designed to keep users engaged. Recent polling bears this out: two-thirds of Australian Gen Z users say they would keep algorithmic recommendations enabled, while nearly half of all Australian adults say they would turn algorithms off entirely.
As the bill moves through Australia’s parliament, the fate of the reform could set a global precedent for how governments hold big tech accountable for algorithmic harm, and whether users will finally get real control over what they see on the social platforms they use every day.