Protect the Children! Australia's Digital Duty of Care and the Expansion of Online Control

There are few phrases in politics more powerful than "protect the children." It immediately establishes the moral terrain. Who, after all, wants children exposed to pornography, grooming, cyberbullying, encouragement of suicide, eating disorders and the other genuine horrors available online?

Australia's proposed Digital Duty of Care begins on precisely this apparently unassailable ground. The Albanese government wants online services to take "reasonable steps" to prevent "foreseeable online harms," shifting regulation away from merely removing objectionable material after something has happened and towards requiring platforms to anticipate and prevent harm in advance.

There is a perfectly respectable argument for doing that. Children should not be deliberately fed pornography by algorithms. Platforms should not knowingly design systems that facilitate grooming. A company discovering that its recommendation system repeatedly directs vulnerable teenagers towards self-harm material should not be able to shrug its corporate shoulders and say that somebody can lodge a complaint afterwards.

The problem begins when we move from those propositions, which almost everybody accepts, to the machinery government proposes to construct around them.

The government's own Digital Duty of Care framework is much broader than simply keeping pornography away from twelve-year-olds. The proposed duty encompasses harms to young people, threats to mental and physical wellbeing, attacks based upon protected characteristics, promotion of harmful practices, threats to national security and "social cohesion," terrorism and other illegal activity. Large or high-risk services would conduct regular risk assessments, while the eSafety Commissioner would be empowered to create mandatory codes dealing with compliance.

That is an enormous conceptual territory.

Words such as "grooming" and "child sexual exploitation" have reasonably definite meanings. "Harm," "wellbeing" and especially "social cohesion" are considerably more elastic.

And once the legal obligation is not merely to remove unlawful material but proactively to prevent foreseeable harm, the incentives facing online platforms change dramatically.

If leaving controversial material online potentially creates regulatory liability while removing it does not, the rational corporate response is obvious: remove first and argue about freedom later.

That is not some fantastic objection dreamed up by opponents of regulation. The United States government has now raised precisely the concern that Australia's proposed framework could encourage over-moderation of lawful speech, particularly where government acquires significant influence over what constitutes foreseeable harm.

A government does not need to establish a Ministry of Truth and employ thousands of censors personally to influence speech. It can create sufficiently broad duties, attach sufficiently serious penalties to them, and allow private companies to perform the censorship themselves as a rational exercise in regulatory risk management.

The government can then say that nobody ordered a particular opinion removed. Technically, that may even be true. The platform simply knew which direction the incentives pointed.

This is where "protect the children" deserves particularly careful scrutiny. Children unquestionably require protection, but invoking them can also make scrutiny of governmental power socially difficult. Once opposition to a regulatory proposal can be rhetorically transformed into indifference towards harmed children, the argument has been half-won before anybody examines what the legislation actually permits.

The Digital Duty of Care is not confined to children. The government's consultation material explicitly describes the objective as preventing serious online harms to Australians, and says responsibility for managing those harms should shift from individuals towards digital platforms.

That is a significant philosophical change. Adults have traditionally been presumed capable of encountering ideas, images and opinions that may disturb, offend or upset them. Liberal societies punish threats, fraud, harassment, incitement and other defined wrongdoing, but they have historically been much more cautious about giving authorities a general responsibility for protecting competent adults from harmful expression.

A duty to prevent "foreseeable harm" potentially moves the boundary. The objection becomes stronger when "social cohesion' enters the vocabulary. Governments inevitably have views about what strengthens or damages social cohesion. So do activists, corporations, academics, religious organisations and political movements. They frequently disagree.

An immigration critic may regard unrestricted discussion of immigration's consequences as essential democratic speech. Another person may regard some of that discussion as contributing to racial hostility and therefore damaging social cohesion. A radical environmentalist may believe disruptive rhetoric is necessary to prevent ecological catastrophe. Somebody else may regard it as encouraging extremism.

During a pandemic, one person's dangerous medical misinformation may be another person's legitimate challenge to official policy. The difficult question is not whether genuinely harmful material exists. Of course it does. The difficult question is who gets to draw the boundary.

Australia already possesses substantial online-safety machinery. Providers are subject to unlawful-material requirements, age-restricted material codes and Basic Online Safety Expectations. eSafety already possesses reporting and enforcement powers, while the new age-restricted material codes include protections concerning pornography and self-harm content.

So before constructing another layer of anticipatory regulation, Australians are entitled to ask a simpler question: how much of the problem involving children should actually be solved in the home?

Here lies an uncomfortable truth that rarely receives the attention it deserves.

For most of human history parents have been expected to supervise the environments their children enter. Nobody would give an eleven-year-old the keys to a car and say, "Drive wherever you like, and government regulators will make sure the roads contain nothing dangerous."

Yet something remarkably close to this has happened with smartphones.

A child can be given a device providing almost unrestricted access to pornography, strangers, addictive games, social-media algorithms and virtually the entire accumulated contents of human civilisation, then disappear into a bedroom with it for six hours.

When the predictable problems arrive, the proposed solution increasingly becomes government regulation. That reverses the natural order of responsibility.

Parents have far more information about an individual child than Canberra ever will. They know whether their thirteen-year-old is mature or vulnerable, whether the child is being bullied, whether sleep is being destroyed by midnight social-media use, whether particular applications are becoming obsessive and whether the child should have unrestricted internet access at all.

There are practical reasons why parental supervision sometimes fails. Many parents are exhausted. Both parents may be working long hours. Some lack technical knowledge. Some underestimate how sophisticated modern platforms have become. Children themselves can circumvent restrictions.

And, uncomfortable though it is to say, some parents simply use the internet as an electronic childminder. None of that means government has no role. Government can prosecute predators. It can require platforms to cooperate against child sexual exploitation. It can make parental controls simple and effective. Schools can teach children how algorithmic manipulation works. Governments can insist upon meaningful transparency about what platforms are doing to children rather than demanding that parents decipher hundreds of pages of terms and conditions.

Those are targeted interventions directed towards identifiable problems. They are quite different from establishing an expanding governmental responsibility to ensure that the online environment is sufficiently safe for everybody.

Indeed, strengthening parental authority could accomplish a substantial part of the child-protection objective without creating an architecture capable of regulating adult political and social discourse.

Make devices genuinely parent-controlled by default. Give parents simple operating-system-level controls over applications, browsing, purchases, messaging and screen time. Make it easy to prevent children installing applications secretly. Give parents understandable activity reports. Provide straightforward filtering tools without requiring a computer-science degree to configure them.

Then insist that parents actually parent. The attraction of the alternative is obvious. Responsibility migrates upwards. Parents hand responsibility to platforms, platforms hand responsibility to regulators, regulators acquire more authority, and politicians announce that another social problem has been addressed. The state becomes the universal parent.

There is a further danger in assuming that this process has a natural stopping point. Regulatory bureaucracies rarely announce that their mission has been accomplished and their powers can now be reduced. New technologies produce new harms, new harms produce new responsibilities, and new responsibilities produce new demands for regulatory authority.

Australia can already see that progression. The Online Safety Act created one framework. Codes and standards followed. Age-assurance requirements followed. The social-media minimum-age regime followed. Now comes a Digital Duty of Care intended eventually to replace existing codes and standards with a broader systems-based model requiring services proactively to manage risks. The government's own impact-analysis material describes precisely that transition. Each individual step can be defended. Taken together, however, they raise a larger question about the kind of internet Australians are constructing.

Do we want an internet on which adults remain presumptively free to communicate, subject to clearly defined laws against particular wrongdoing? Or do we want an internet increasingly engineered around a governmental conception of safety, where platforms must continuously anticipate what regulators might subsequently decide was a foreseeable harm?

Children should not become rhetorical pawns in a much larger transfer of authority over online life. If the problem is children encountering material inappropriate for children, then begin with measures directed towards children, parents and clearly identified harmful practices.

If government instead seeks powers affecting what adults can encounter and say, those powers should be defended openly on their own merits. There is a legitimate place for law in cyberspace, just as there is in the physical world. Predators should not be protected by slogans about free speech. Companies should not be allowed knowingly to engineer dangerous products for children and then disclaim all responsibility for the consequences. But neither should genuine concern for children become a blank cheque for an ever-expanding system of digital supervision.

https://networkaffects.substack.com/p/digital-duty-of-care-australias-new