A Censorship Machine Is In This Bill!

The Albanese Government is building a censorship machine, and the most important thing to understand is that it doesn't look anything like the censorship machine most people have in their heads.

There's no Department of Censorship. There's no miserable bureaucrat sitting in Canberra reading your Facebook posts and stamping APPROVED or BANNED across them. Anthony Albanese doesn't need to telephone Mark Zuckerberg and dictate which political opinions Australians are permitted to read. That would be crude, obvious, and politically explosive.

This is much smarter. The Government's proposed Digital Duty of Care creates a legal concept called a “safe online environment”. It requires online services to protect Australians from specified categories of harm and makes those companies legally responsible for managing the risks.

That sounds perfectly reasonable until you ask the question sitting underneath the entire scheme: Who decides what is harmful?

Start with proposed section 25C of the Online Safety Act, which would define “seriously harmful material and conduct”. A lot of what's there is exactly what you'd expect: child sexual abuse and exploitation, grooming, material encouraging sexual or extreme violence, explicit threats of rape or death, material encouraging suicide or self-harm, terrorist material, and material encouraging criminal offences. All that is well enough, even if its already covered by existing law which makes it illegal.

If the legislation stopped with clearly defined criminal and seriously dangerous conduct, this would be a very different debate. But it doesn't. Keep reading until paragraph 25C(1)(m). There, you find a catch-all covering “any other material or conduct” determined under the next subsection.

Then read subsection 25C(2). The Minister may, by legislative instrument, determine additional material or conduct that the Minister is satisfied “may cause serious harm”.

There it is. Parliament isn't creating a single closed list and locking it into an Act. It is building an expandable system in which additional categories can be fed into the machinery later.

And notice what the test isn't. It doesn't say the additional material must first be made criminal by Parliament. It doesn't say a court must first declare it unlawful. The Minister must be satisfied that it “may cause serious harm”.

Think about how much can eventually be squeezed into a concept like that. We are talking about the internet, the principal communications system of modern Australia, and Canberra is proposing legislation under which a minister can expand categories of regulated harmful material through delegated legislation. And we're supposed to be reassured that the power will be used responsibly. Right.

The politician administering this law, Communications Minister Anika Wells, is bad enough. But there could be worse. Wells will eventually leave office. Governments will change. Ministers will change. Political fashions will change. Ideas regarded as ordinary today may be treated as socially dangerous 20 years from now. Politicians change. Powers remain.

Then you reach proposed section 25D, and the problem becomes even clearer. This section deals specifically with material and conduct considered harmful to children. Again, some categories are unsurprising: pornography, material encouraging disordered eating, abuse, harassment, and bullying.

Then paragraph 25D(1)(c) captures material or conduct that “encourages or promotes hostile attitudes towards women or gender equality”. Huh?

We're no longer simply talking about child pornography, grooming, threats, or criminal conduct. We're talking about attitudes. What exactly constitutes a “hostile attitude” towards gender equality?

Suppose somebody publishes a blistering attack on modern feminism. Suppose a Christian preacher teaches that husbands and wives have different responsibilities within marriage. Suppose somebody campaigns against gender quotas. Suppose somebody argues that biological males shouldn't compete in women's sport. Which of those crosses the line?

Who knows? But that's precisely the point. The boundary is no longer something obvious like “don't threaten to kill somebody”. We're entering territory involving social attitudes and ideological judgement. And once government moves from policing objectively unlawful conduct into regulating broader concepts involving attitudes and harm, politics inevitably enters the equation.

Section 25D also contains further catch-all provisions dealing with other material or conduct capable of inflicting serious harm on a child, along with another power for the Minister to determine additional categories. There it is again. The system can grow.

If you think that concern is some paranoid fantasy, keep reading until proposed section 27A. The legislation expressly authorises the eSafety Commissioner to advise the Minister about “emerging categories of harm arising from the use of online services”. Those words deserve attention: emerging categories of harm.

The Commissioner can also advise the Minister about the need to make or amend legislative instruments under this part of the Act. Then section 27B helps connect these provisions together. Before exercising certain powers to make legislative instruments, the Minister must seek advice from the totally censorious and politically compromised eSafety Commissioner. Those powers include ones dealing with additional categories of seriously harmful material, additional material harmful to children, and additional design features deemed to have negative behavioural impacts.

Now stand back and look at what they've constructed. The legislation establishes categories of harm. The regulator watches for emerging categories of harm. The regulator advises the Minister. The Minister has mechanisms for expanding parts of the regulatory system.

That is a regulatory ratchet. Today, the categories look one way. Tomorrow, some new social panic erupts. A new form of supposed harm is identified. The bureaucracy examines it. The Minister responds. Another category enters the system.

And every turn of that ratchet will have its own reasonable-sounding explanation. It's only a minor amendment. It's only about safety. It's only addressing a new threat. It's only protecting vulnerable people. Won't somebody think of the children! Click. Then another problem emerges. Click. Then another.

Nobody announces the creation of a censorship regime with a brass band marching down Commonwealth Avenue. Government power usually grows by accumulation, one apparently sensible expansion at a time. And, as it grows, it eats into freedom, particularly freedom of speech which is the cornerstone of any true democracy.

For years, we've watched political disagreement increasingly described using the vocabulary of danger: misinformation, disinformation, extremism, radicalisation, division, unsafe speech, hate, and threats to social cohesion. During COVID, people challenging lockdowns and vaccine mandates weren't merely accused of being wrong. Their speech was described as dangerous and harmful.

In the gender debate, traditional Christian teaching is routinely accused of causing harm. Women objecting to biological males entering female sports are accused of making others unsafe. Arguments over immigration quickly become arguments about extremism, division, and social cohesion. Politics itself is increasingly translated into the language of harm.

And now Canberra wants an expandable online regulatory system built around that very concept. The Government will insist that it isn't proposing to censor political opinions. Indeed, “misinformation” isn't currently sitting among the listed harms in section 25C. But that doesn't settle the argument.

The question isn't merely what Anthony Albanese's Government says it intends to regulate today. The question is what machinery is being created for tomorrow. If the Government wanted a tightly confined law targeting child exploitation, threats, terrorist recruitment, and other objectively defined horrors, it could write one even though it wouldn't need to, given existing laws already cover all of that.

Instead, it has proposed a framework containing expandable categories of harm, ministerial powers, and an explicit statutory role for identifying “emerging categories of harm”. That's the real story in all of this. And that's what should be at the forefront of the debate.

Because once the machine exists, the political fight changes. Future governments no longer have to construct the framework from scratch. They argue about what else should be put inside it.

So forget whether you trust Anthony Albanese. Imagine the Australian politician you distrust most. Make that person Prime Minister. Give their party government. Give them a Communications Minister. Let their government decide what new forms of online harm demand action within the powers Parliament has provided.

Still comfortable? That's the test. Never judge a government power according to the politician asking for it. Judge it according to the politician you fear may eventually inherit it.

And when the power concerns the flow of information itself, Australians should be especially reluctant to hand over the keys.

“The natural progress of things is for liberty to yield, and government to gain ground.”
– Thomas Jefferson