Australia is on the verge of making another major move in the regulation of social media sites with the federal government planning to introduce a change in the form of a Digital Duty of Care on digital platforms.

One of the main components of the proposed framework will enable social media users to have more control over the content shown to them as suggested by their algorithm, and even an option to disable personalised suggestions. The government has indicated that the bigger picture is to hold online platforms more accountable to detect and avert foreseeable online harms instead of leaving it mainly on the user to guard oneself.

The offer comes at a time when Australia is working to enhance its online-safety system, which has already established minimum-age regulations of social media. Nevertheless, the Digital Duty of Care is a suggested legislative change. As of 7 September 2026, the government’s Digital Duty of Care legislation has not yet been introduced to Parliament. The government is expected to release an exposure draft and introduce the bill this week. 

What Is Australia’s New Social Media Algorithm Proposal?

The proposed Digital Duty of Care by the Australian government would place greater liability on online safety on the hands of digital platforms.

The policy can be traced back to the stand-alone 2024 review of the Online Safety Act 2021 in Australia by Delia Rickard. It made 67 recommendations, among which is the establishment of a statutory duty of care to ensure that online services are more proactive and systems-based in the approach to harm prevention. The government then undertook to enact a Digital Duty of Care.

That bigger structure has now been associated by the government with how recommendation algorithms shape what is shown to the users.

Through the proposal that the government is preparing, social media platforms will be obliged to offer the users a tool to unsubscribe to the algorithmically suggested content. According to recent government claims and reportage, users might be offered an option of either continuing to be fed algorithmic-curated content and seeing content based more directly on accounts that they follow. The exact functioning of that feature will be based on the final legislation and regulations.

This is significant since the proposal is not merely on deleting individual posts. It targets the mechanisms that dictate the content that is prioritised and shown to users.

How Would an Algorithm Off-Switch Work?

The idea of social media recommendation systems is to select, rank and personalise the content.

These systems may take into account what one views, likes, shares or searches, what accounts they follow, the duration of interaction with specific content and other data relating to their account, depending on the platform.

These signals are then used by the system to decide which posts, videos or other contents are likely to be of interest to the person.

This will be able to personalise feeds and assist users in finding out material that is not necessarily in the feeds of accounts they already follow. Nevertheless, eSafety has also determined possible risks whereby the recommender systems enhance harmful content on numerous occasions or add to the cumulative effects of harmful content. The regulator indicates that recommender systems can have both positive and negative impacts and has urged more robust user controls and safer design.

A user would have a further control level with an algorithm off-switch.

Provided it is implemented as is being reported, a user who opts to turn off personalised recommendations might be presented with a less personalised experience, perhaps with a greater focus on content posted by accounts that the user has chosen to follow. The final legislation and implementation of the particular platform would dictate the exact design of the feed.

The notion also cannot be mixed with shutting down the entire algorithmic infrastructure of a platform. Recommendation systems are capable of numerous tasks, such as ranking, moderation and safety related tasks. The given measure is primarily interested in algorithmically suggested material that is offered to users.

Why Is Australia Targeting Social Media Algorithms?

The priorities of the government indicate a more comprehensive change in policy regarding online-safety, no longer focusing on the content but on the design and functioning of online platforms.

A Digital Duty of Care would entail platforms to undertake reasonable efforts to detect and mitigate predictable risks. The review of the Online Safety Act (2024) claimed that such an approach would resolve such harms not only caused by particular content but also by platform design, recommendation systems and other technological characteristics.

eSafety’s May 2026 position statement similarly highlighted the importance of recommender systems in determining users’ online experiences. It stated safer design must have more control over algorithm settings by users and more transparency on the design and use of recommendation systems.

Children and young people are of particular interest in the policy debate. Australia already has a social-media minimum-age system in operation in which age-restricted sites are obligated to make reasonable efforts to ensure that Australians under 16 do not have accounts.

In March 2026, the rules were updated to include services with account-based recommender systems and features such as endless feeds, feedback mechanisms and time-limited content. 

The broader policy discussion is increasingly focused not only on who can access social media, but also on how platforms design recommendation systems and manage risks once users are on their services. 

What Would the Digital Duty of Care Require From Big Tech?

This is because the ultimate requirements will not be evident until the government releases its draft legislation which will then go through Parliament.

Nevertheless, it is anticipated that the policy structure will put a proactive burden on the digital platforms to detect and prevent foreseeable harms as opposed to reacting after harmful contents have already inflicted harm.

The Digital Duty of Care as the current commitment of the government is defined as a framework that would entail the digital platforms to adopt reasonable measures to avert foreseeable online harms. Risk assessment and risk mitigation are also the main parts of the model that the independent review that gave rise to the proposal predicted.

Recent coverage states that the suggested bill would not only deal with pernicious or unlawful material but also provide further safeguards to kids, such as dangers linked to features like bullying and content relating to body-image. This information is informed by the coverage of the draft proposal put forward by the government and cannot be considered as the final legislative provisions before the exposure draft and bill are officially published.

The proposed proposal also entails major financial fines for non-adherence. According to ABC, businesses might pay a fine in excess of A$100 million as part of the proposed government framework. The exact provisions of penalties must however be verified with the final bill introduced.

The difference is significant since the current social-media minimum-age system in Australia already has its punishment system. In the current laws, social media providing age-restricted services can attract fines of up to A$49.5 million because of the inability to take reasonable actions to avoid under-16s holding accounts. Another government bill awaiting Senate approval seeks to add penalties to those.

What Role Could the eSafety Commissioner Play?

The Safety Commissioner is the independent online-safety regulator in Australia, and already has considerable regulatory duties under the Online Safety Act 2021.

Within the current social-media minimum-age system, eSafety will be tasked with evaluating compliance and deciding whether platforms have made reasonable efforts to ensure under-16s do not maintain accounts. The Online Safety Act also gives the regulator its current investigative and enforcement powers.

The Digital Duty of Care is supposed to enhance the wider regulatory framework, however, the specific enforcement authorities that would be applied to the new obligations will be dependent on the legislation.

The government has mentioned that the duty of care will place the legal responsibility of digital platforms themselves. eSafety has also indicated that it would like the future framework to establish strong and enforceable requirements covering online harms, such as those related to recommender systems and algorithms.

This is not then a mere suggestion to have the regulator have increased powers to delete individual posts. The wider policy trend is the systematic regulation of the manner in which platforms detect, avert and deal with risks.

Opt-Out vs Opt-In: Why Is This Debate Important?

Among the key political issues is whether algorithmic personalisation should still be on default, or users must be allowed to switch it off, or whether personalisation should be turned off unless a user deliberately opts to personalise.

An opt-out model implies that personalized recommendations would be active, unless the user decides to switch them off.

A default would be reversed by an opt-in model. Algorithms personalisation would require users to take active decisions before it could be turned on.

Recently the government has been talking about an opt-out model. Attorney-General Michelle Rowland noted that the government was weighing whether or not the mechanism would eventually be opt-in or opt-out, and recent coverage has suggested that the draft government proposal will adopt an opt-out system.

The Greens in Australia have campaigned in support of a more vigorous opt-in policy. The Online Safety Amendment (Fix Our Feeds) Bill 2026, a private bill sponsored by Senator Sarah Hanson-Young, would require social media services to give users the option to opt out of receiving algorithmically recommended content. The Greens have also advocated a stronger opt-in approach, under which users would actively choose to enable algorithmic personalisation rather than having it enabled by default.

This distinction is also considerable since the default setting will define the extent of effort that users will need to exert in order to alter their experience.

What Are Critics Saying?

The proposal has also created an issue of government power, censorship and freedom of expression.

Critics of the government have suggested that increasing the level of government control over how content is handled by platforms would pose a threat to free speech. Representatives of the coalition have indicated that they desire to view the fine-tuning legislation prior to making a decision as to whether the government has achieved the right balance between online security and individual rights.

These issues do not imply that an algorithm-control mechanism would necessarily involve censorship. The offered off-switch is more of a suggestion regarding content in terms of the manner of recommendations, but not the government having the ultimate power to determine what each person is allowed to post or read.

Implementation practical questions also exist. A non-personalised feed might appear different on different platforms, and users might not always realise the difference between algorithmic recommendation, chronological ranking, moderation and other types of automated content selection.

Another argument by those who support stricter regulation is that an opt-out system might be inadequate. Other campaigners and specialists have insisted that it should be the users, especially young or vulnerable, who must put up with harmful recommendation systems, and disable them themselves.

This is the reason why the opt-in and opt-out question is a significant element of the political discussion.

What Is the Current Status of the Proposal?

By 7 September 2026 the Digital Duty of Care legislation by the government was yet to be enacted and the bill by the government was yet to be presented in the Parliament.

The government will issue an exposure draft and present the legislation when the current parliamentary sitting is underway. Before the draft and final bill is published, critical information such as the specific algorithm controls, scope of covered services, enforcement provisions and penalty provisions are subject to change.

As a direct result of the independent Online Safety Act review, the government is already pledged to enact legislation of a Digital Duty of Care. In February 2025, the review was tabled in Parliament, following 67 recommendations to enhance the online-safety framework in Australia.

Once the government presents the bill, it would then have to undergo the process of the Australian parliament. That usually involves a consideration and debate in the House of Representatives and Senate prior to a bill becoming law followed by Royal Assent.

The proposal should not be a new obligation but a legislative proposal until those steps are made.

How Does the Fix Our Feeds Bill Differ?

Another legislative effort is the Online Safety Amendment (Fix Our Feeds) Bill 2026.

It is not government legislation but is a privately sponsored bill by Greens Senator Sarah Hanson-Young. It was presented on 1 April 2026 in the Senate and is still pending before the Senate in the list of the Australian Parliament.

The bill would revise the Online Safety Act 2021 to compel social media services to offer a choice of not receiving algorithmically-suggested material to users. It also includes suggested reporting, risk-assessment and risk-mitigation requirements on the online service providers.

The Greens have taken the bill to push towards a more aggressive stance to the regulation of algorithms, such as an opt-in model.

It cannot then be referred to as the Digital Duty of Care bill of the government. The two initiatives handle similar issues but are sponsored by different people, through legislation and the approaches that are proposed.

What Could This Mean for Social Media Users?

Should the proposal of the government turn into a legislative act in its substantially reported version, Australian users might have more control over the curation of their social media feeds.

The current algorithm-based experience might be kept by users who would like to obtain personalized suggestions.

To users who prefer not to use recommendations, the site can instead give them a less personalized feed that is more reliant on accounts or content they have explicitly followed. Precisely, the experience would be determined by the ultimate legislation and the manner in which every covered platform does the requirements.

The transformation may also render recommendation settings a more apparent aspect of user-platform relationships.

In the case of children and young people, the implications are more extensive. The current framework of minimum-age in Australia already considers recommender systems and engagement-oriented features in the regulatory definition of the age-restricted social media. The Digital Duty of Care would widen the policy discussion beyond just the debate over who is allowed to access an account, to a consideration of how platforms themselves handle foreseeable risks. 

Why This Matters for the Global Big Tech Debate

Australia’s proposal is part of a much wider international discussion about how governments should regulate large digital platforms.

The central question is increasingly moving beyond whether a platform should remove individual pieces of harmful content. Regulators are also examining how platform architecture, recommendation systems, data use and engagement mechanisms can influence what people encounter online.

Australia’s Digital Duty of Care approach is informed by the country’s own Online Safety Act review and sits alongside regulatory developments in other jurisdictions, including the European Union’s Digital Services Act and the UK’s Online Safety Act. Australia’s eSafety Commissioner has also identified recommender systems as an area requiring greater attention to safety, transparency and user control.

That does not mean Australia’s proposal will automatically establish a global standard. Its significance lies instead in adding another major regulatory experiment to the international debate over how governments should hold large technology platforms accountable. 

Final Thoughts

Australia’s proposed Digital Duty of Care represents a shift towards making social media companies more directly responsible for the risks created by the systems they design and operate.

The possibility of allowing users to turn off algorithmically recommended content is one of the most visible elements of the proposal. But the wider reform is about more than an algorithm switch. It is intended to move online safety towards a preventative model in which platforms are expected to identify risks, assess their impact and take steps to reduce foreseeable harm.

The proposal is not yet law. The government is expected to release its draft legislation this week, after which the details will face parliamentary scrutiny and political debate.

The most important issues to watch will be the final design of the algorithm-control mechanism, whether the government maintains an opt-out default or moves towards an opt-in system, the powers given to regulators, the penalties for non-compliance and how broadly the new obligations apply to major digital platforms.

For Australia’s social media users, the outcome could determine how much control they have over one of the most influential parts of the modern online experience: the system deciding what appears in their feed.