Australia Drafts a Digital Duty of Care That Would Regulate Social Media Algorithms

Australia's Labor government is preparing draft legislation that would impose a "digital duty of care" on social media platforms, requiring them to protect users from harmful content or face fines of more than $100 million. The Guardian reported on September 5 that Communications Minister Anika Wells is expected to release the draft as federal parliament returns this week, and that Prime Minister Anthony Albanese is planning to promote the proposal at the United Nations General Assembly in New York later this month, an unusual piece of stagecraft that signals the government sees the issue as a global cause rather than a domestic regulatory matter.

The proposal is the latest chapter in Australia's aggressive approach to online safety, a record that already includes the world's first under-16 social media ban, which took effect in late 2025 and requires platforms to verify users' ages. The digital duty of care would go beyond that ban by regulating how platforms rank and recommend content, not just who is allowed on them. Under the draft, popular platforms would have to let users switch off content-ranking algorithms, and companies would be required to better identify problematic or illegal content, with the aim of protecting vulnerable users, particularly children.

The political battle lines are already drawn. Opposition Leader Angus Taylor immediately called the plan an attempt to censor the internet, and the Coalition is expected to oppose it in the Senate, where the government will likely need the Greens to pass the legislation. The debate echoes the fate of an earlier anti-misinformation bill, which the government abandoned after similar opposition, and the outcome of this proposal will test whether Canberra can regulate the algorithms that shape what Australians see online.

Key Facts

The draft's core requirement is algorithmic transparency with teeth. The Guardian reported on September 5 that platforms covered by the duty would be required to offer users a way to switch off content-ranking algorithms, giving Australians the option of seeing posts in chronological order rather than in the engagement-optimized order the platforms choose. The same legislation would oblige platforms to identify and act on problematic and illegal content more effectively, with a stated focus on protecting children and other vulnerable users.

The enforcement mechanism is designed to be meaningful. The Guardian reported on September 5 that the draft includes fines of more than $100 million for breaches, a figure intended to exceed the annual profit that a platform might earn from ignoring the rules, and that it would grant the eSafety Commissioner new investigation powers along with expanded access for independent researchers. The researcher access is a notable addition, because academics studying platform harm have long complained that they cannot get the data they need to document what the algorithms actually do.

The political context is the recent history of Australian platform regulation. The Guardian reported on September 5 that Angus Taylor described the proposal as an attempt to "censor the internet," and his shadow minister for home affairs, Jonno Duniam, compared it to the anti-misinformation bill that the government abandoned after a similar backlash. The Greens are likely to be pivotal in the Senate, since the Coalition and One Nation are expected to oppose the bill, and the government will need their votes to pass it into law.

The government's international framing is deliberate. Albanese is expected to promote the duty of care at the UN General Assembly later this month, presenting Australia's approach as a model for other countries wrestling with the same problems. The under-16 ban, which preceded this proposal, was itself a global first, and Canberra has shown a willingness to move before its allies, even when that means navigating criticism from the platforms and from civil-liberties groups who argue the measures go too far.

Analysis

What this really means is that Australia has decided that the harms of social media are a product-safety problem, not a free-speech problem, and the duty of care is the regulatory framework that follows from that decision. The concept borrows from tort law, where a duty of care requires a business to take reasonable steps to avoid foreseeable harm, and applying it to platforms is a way of shifting the burden of proof: instead of regulators having to show that a specific post caused harm, the platform must show that its systems were reasonably designed to prevent harm in the first place.

The bigger picture here is that the algorithm is becoming the new frontier of platform regulation, and Australia is trying to cross it before anyone else has worked out how. The under-16 ban was comparatively simple because it regulated access; the duty of care regulates the recommendation systems that determine what millions of people see every day, and that is a far more intrusive form of control. Taylor's accusation of censorship is not entirely without force, because telling a platform how to rank content is closer to editing than to age verification, but the government's answer, that users should at least have the choice to turn the algorithm off, is a narrower and more defensible intervention than a ban on specific content.

The 4.4 percent figure that Taylor cited is the key to understanding the politics. The opposition leader dismissed the under-16 ban's effect as immaterial, citing a roughly 4.4 percent reduction in teen usage, but the government's response is that the ban was never meant to be the whole answer. The duty of care is the follow-through: age bans change who is on the platform, while algorithmic rules change what everyone, including adults, is exposed to. Whether that distinction survives contact with the platforms' legal teams, which will argue that ranking systems are protected expression and that a duty of care is impossibly vague, is the central question of the coming legislative fight.

Why It Matters

For Australians, the duty of care would change the daily experience of social media if it passes, giving every user the option to see posts without algorithmic ranking and creating a legal backstop when platforms fail to remove dangerous content. For the platforms themselves, the stakes are enormous: a $100 million fine is large enough to change behavior, but the real cost would be the precedent of a major Western democracy regulating recommendation algorithms, a move that could ripple through Canada, the United Kingdom and the European Union, all of which are watching Australia's experiment with platform regulation closely.

For the global debate over online safety, the proposal is a test of whether algorithmic regulation is politically survivable. Australia has already shown that age bans can be enacted and defended, but the duty of care targets the engagement-optimization systems that are the platforms' core business model, and the platforms will fight it harder than they fought the age ban. If Australia passes this law, it will give regulators everywhere a template; if it fails, it will be cited for years as proof that algorithmic transparency is a bridge too far.

Next Up

The first milestone is the release of the draft legislation this week, which will reveal the precise scope of the duty, including which platforms are covered, what counts as harmful content and how the algorithm opt-out will work in practice. Watch for the consultation process to draw submissions from the platforms, child-safety advocates and civil-liberties groups, and for the Greens to signal whether the government's draft goes far enough to win their support in the Senate. Albanese's UN General Assembly appearance later this month will be the first test of whether the government can convert its domestic proposal into international influence, and the reaction of other leaders will show how much appetite there is beyond Australia for regulating the algorithms that shape the internet.