The Albanese government wants to make the internet safer. At least, that's how it's selling its latest legislation.
The Online Safety Amendment (Digital Duty of Care) Bill 2026 promises to protect children, tackle online abuse and hold technology companies accountable.
Those sound like reasonable goals. Few Australians would object to stopping child exploitation, violent threats or the spread of intimate images without consent.
But buried beneath the promises of online safety is a major expansion of government regulation over digital platforms, including new investigative powers for the eSafety Commissioner.
The legislation passed the House of Representatives on 28 October and now faces the Senate.
Australians should be paying attention.
What exactly is the Digital Duty of Care?
The proposed law would require online services to identify potential risks and take appropriate steps to prevent serious harm before it happens.
It covers far more than Facebook or X. Online games, websites, messaging services, search engines and AI chatbots could all fall within its scope.
Here's a simple breakdown of the main rules:
Online platforms must prevent serious harm. Companies would need to assess risks and introduce safeguards against specified material and conduct, including child exploitation, violent threats, terrorism-related material and certain forms of harassment.
Stronger protections for under-18s. Platforms must address risks involving pornography, bullying, eating disorders, dangerous challenges and content promoting hostility towards women or gender equality. The House also added protections concerning AI chatbots encouraging emotional dependence in children.
More control over algorithms. Australians aged 16 and over would be able to choose a feed based on accounts they follow rather than algorithmically recommended content.
Faster complaint responses. Certain cyberbullying and adult cyber abuse complaints would require a response within 24 hours instead of 48.
Expanded eSafety powers. The regulator would gain additional investigative, information-gathering and enforcement powers, including the use of fictitious online identities.
Massive financial penalties. Providers breaching the core duty could face fines of up to $109.2 million.
Some of these proposals have obvious benefits. Giving people more control over algorithms, for example, is hardly an attack on free speech.
The trouble lies in how the broader powers might be interpreted and enforced.
Government Sock Puppet Accounts? Yes, They're in the Bill
One particularly controversial provision allows the eSafety Commissioner and approved researchers to operate what the legislation calls "sock puppet identities".
In plain English, these are fake online identities used to investigate and test digital services.
The eSafety Commissioner would be able to use fictitious accounts to observe platform behavior, test safeguards and generate lawful material, including with artificial intelligence.
Approved university researchers would also be permitted to use such accounts under specific conditions and ethical requirements.
Communications Minister Anika Wells argues the accounts are necessary to test whether technology companies are following the rules. For example, investigators could check whether an underage user can bypass a platform's age restrictions.
There is a legitimate purpose for that kind of testing.
However, the idea of a government regulator operating undercover accounts online deserves scrutiny, especially when the same regulator has substantial powers to investigate and penalize platforms.
What information will these accounts collect? How will their use be monitored? What safeguards will prevent the powers from being misused?
The bill requires eSafety to publish an annual statement about its sock puppet activities. That provides some transparency, but it does not eliminate concerns about oversight.
One claim circulating online also needs correcting. The bill does not ban ordinary Australians from having anonymous social media accounts. The sock puppet provisions authorize particular investigative activities. They do not introduce a general requirement for everyone else to obtain government permission to remain anonymous.
The genuine concern is the expansion of undercover regulatory powers, not an invented ban on anonymous users.
The Free Speech Problem
Perhaps the greatest concern is how the legislation deals with harmful content.
For adults, the bill specifies categories of serious harm, including threats of violence, sexual exploitation and certain criminal conduct.
For children, the definition is broader.
It includes material or conduct that may cause serious harm, with examples involving hostility towards women or gender equality, bullying and seriously harmful practices.
Preventing genuine abuse is one thing. Determining where legitimate disagreement ends and harmful expression begins is considerably more complicated.
Consider a heated political discussion about gender equality, social policy or cultural issues.
A strongly expressed opinion is not automatically harassment, hatred or a threat. Yet platforms operating under broad safety obligations may decide that removing controversial discussions is easier than defending them.
With penalties potentially reaching $109.2 million, companies have a powerful incentive to avoid regulatory trouble.
They do not necessarily need an explicit government order to remove lawful content. The threat of substantial financial consequences may encourage them to restrict material themselves.
That is how legitimate expression could be caught in the process, even when lawmakers insist censorship is not the intention.
The Australian Human Rights Commission has also warned that overly broad online safety rules risk restricting freedom of expression and access to information.
Those concerns deserve more than political assurances.
Does It Violate Freedom of Speech?
Australia does not have the equivalent of America's First Amendment.
The Australian Constitution contains an implied freedom of political communication, recognized by the High Court. It limits government power rather than providing an unrestricted personal right to say whatever one wishes.
Whether the Digital Duty of Care would breach that constitutional protection is a legal question that cannot be answered simply by calling the legislation censorship.
However, a law does not need to be declared unconstitutional to create problems for free expression.
If platforms begin restricting otherwise lawful political commentary because they fear regulatory penalties, Australians could experience less freedom to participate in public debate.
The government has included exemptions for professional news, artistic material, education and other public-interest content.
That is welcome, but it does not settle how ordinary users, independent commentators and smaller publishers will be treated when platforms make moderation decisions.
Freedom of expression should not depend on whether a technology company's moderation system considers an opinion worth the regulatory risk.
Is Labor on a Censorship Mission?
There is a reason critics are suspicious of the Albanese government's growing interest in policing online activity.
In 2024, Labor pursued its controversial misinformation and disinformation legislation, which would have given the Australian Communications and Media Authority additional oversight powers over digital platforms.
The proposal was abandoned after failing to secure enough Senate support.
The government also introduced Australia's under-16 social media restrictions, which came into effect in December 2025.
Now comes the Digital Duty of Care, extending regulation across a much broader range of internet services.
Each policy has its own stated purpose, but together they represent a significant expansion of government involvement in Australians' digital lives.
Labor denies the latest bill is designed to regulate political opinions or resurrect its failed misinformation legislation.
Importantly, the government also removed an earlier proposal that would have allowed the minister to add new categories of serious harm without passing another Act of Parliament.
That change addressed one of the strongest criticisms of the original draft.
Nevertheless, the bill still leaves considerable responsibility and discretion with the eSafety Commissioner and the companies expected to enforce these obligations.
The concern is not that Labor has already outlawed criticism of the government. It hasn't.
The concern is that successive governments are becoming increasingly comfortable with giving regulators greater influence over the rules governing online expression.
The more authority governments accumulate over the digital public square, the more important it becomes to question how that authority could be used by future governments.
Safety Shouldn't Come at the Cost of Free Speech
Children deserve protection from exploitation, predators and genuinely dangerous online activity.
Technology companies should face consequences when they knowingly ignore serious risks.
But those objectives do not justify giving regulators powers without clear limits, meaningful oversight and protections for lawful expression.
The Digital Duty of Care contains worthwhile measures. It also contains provisions that could encourage excessive content removal, place significant compliance burdens on online services and expand the government's ability to investigate the internet.
Australians should demand clear definitions of harm, stronger independent oversight of covert investigations, reliable appeal mechanisms and explicit protections for lawful political debate.
The Albanese government says it wants to make the internet safer.
It now needs to demonstrate that Australians will remain free to question, criticize, offend and disagree without their lawful opinions becoming casualties of its online safety agenda.
This fits a pattern. The same government previously pursued misinformation and disinformation laws that collapsed under free-speech objections. It introduced a social media age ban for under-16s. It has steadily expanded the eSafety Commissioner’s reach. Vague standards plus heavy penalties plus an empowered regulator produce a system in which platforms become the enforcement arm of government preferences. Content that challenges prevailing views is more likely to be restricted than content that aligns with them. That is not neutral safety regulation. It is a structural preference for control.
Parents already have tools and responsibilities for what their children see. Criminal law already covers genuine threats, child abuse material, and terrorism. A duty of care framed this way does not simply fill gaps. It transfers editorial power from individuals and companies to regulators and risk-averse compliance teams. The result is less open discussion, more pre-emptive removal, and greater government influence over what Australians can say and read online.
At the end of the day, the concern being raised by Australians online is simple. The government has a responsibility to protect people from real harm, but it has no business deciding what Australians can think, say or read simply because someone might find it offensive, or because the government doesn't like being criticized.
0 Comments