Overview

I support effective measures to address unlawful online conduct and to give Australians greater control over the technologies that shape their online experience. I also support appropriate measures that empower parents to manage their children’s online safety.

However, I do not support an online-safety framework that allows lawful information or expression to be suppressed because government, a regulator or a digital service considers it harmful.

The distinction between lawful and unlawful conduct should remain meaningful.

The Digital Duty of Care should be based on a straightforward approach: protect Australians from unlawful conduct, give individuals genuine control over their online experience, empower parents to protect their children, and strengthen Australians’ ability to navigate online environments independently.

Where a particular risk arises from the operation of technology, regulation should address that technology. Where conduct is unlawful, the law should address the conduct. Where information and expression remain lawful, individual choice should remain the starting point.

In my view, online-safety policy should:

protect Australians from unlawful conduct and direct threats to their rights and safety;
empower Australians with genuine control over their online experience;
empower parents with effective tools to protect and guide their children;
uplift Australians through transparency, information and digital literacy; and
control only where genuinely necessary and authorised by law.

The Digital Duty of Care should be amended to reflect these principles.

Online-safety policy should protect, empower and uplift Australians. Control should be reserved for circumstances where it is genuinely necessary and authorised by law.

1. Preserve the distinction between lawful and unlawful conduct

Australian law already provides substantial protections against serious unlawful online conduct. Existing arrangements address child sexual exploitation material, image-based abuse including fake intimate images, serious adult cyber-abuse, cyberbullying of children, terrorist and abhorrent violent material, and other illegal and restricted online content.

The Digital Duty of Care is therefore being introduced into an already extensive regulatory framework.

Where gaps remain in protection against unlawful conduct, those gaps should be identified and addressed directly. A preventative duty may also appropriately require digital services to provide effective safety features, complaint mechanisms and user controls.

It should not create a parallel system through which lawful information can nevertheless be effectively suppressed because it is considered harmful.

If government considers particular information or conduct sufficiently dangerous that Australians should no longer be permitted to access or communicate it, the appropriate course is to put that proposition to Parliament.

That process requires government to identify the proposed restriction, justify it publicly and secure parliamentary approval.

The Digital Duty of Care should not achieve substantially the same outcome indirectly by requiring technology companies to suppress information that Parliament has left lawful.

If Parliament has left information lawful, Australians should remain free to access it. If government wants to change that position, government should have to persuade Parliament to change the law.

2. Harm alone is not a sufficient basis for restricting lawful information

Information can be lawful and harmful at the same time. That does not mean the possibility of harm is irrelevant. It does mean that harm alone does not establish who should have authority to decide whether another person may encounter the information.

Australians routinely encounter confronting, offensive, incorrect, extreme, distressing and potentially harmful ideas through books, newspapers, magazines, films, political debate, religious organisations, universities and ordinary conversation.

Government does not generally respond to these risks by determining which lawful ideas individuals are permitted to encounter.

The same starting principle should apply online.

This does not prevent intervention against threats, exploitation, unlawful harassment, incitement or other unlawful conduct. Nor does it prevent Parliament from deciding that particular material requires specific legal restrictions.

What it should prevent is lawful information acquiring the practical consequences of illegality through regulatory action without the law itself being changed.

Being protected from making a decision is not the same as being empowered to make one.

The Digital Duty of Care should protect Australians from the unlawful conduct of others without unnecessarily assuming responsibility for Australians’ own decisions about lawful information.

3. Regulation should address what is genuinely different about the online environment

There are legitimate reasons for specific regulation of digital services.

Recommendation systems can continuously select material based on observed behaviour. Digital services operate at enormous scale and speed, collect substantial information about users, optimise systems for engagement and can repeatedly place material before users without them actively seeking it.

These characteristics can create risks that do not have straightforward offline equivalents.

Regulation should be directed at the feature creating the distinctive online risk.

Concerns arising from recommendation systems should be addressed through meaningful user control over those systems. Concerns about data collection should be addressed through appropriate privacy and data protections. Concerns about platform design should be addressed through targeted design requirements.

These issues do not, by themselves, establish a need to restrict the lawful information being communicated.

Where government would not restrict an Australian’s access to the same lawful information in a book, newspaper, magazine, film or conversation, the additional technological risk should be identified and addressed rather than the underlying information.

This distinction allows government to regulate genuinely novel features of the online environment without assuming a broader power over lawful expression.

4. User control should be the primary response to recommendation systems

I support the principle underlying My Feed, My Way. Giving Australians meaningful control over whether they use algorithmically recommended feeds is an appropriate response to one of the genuinely distinctive features of the online environment.

This approach should extend throughout the Digital Duty of Care.

Users should have simple and effective ways to opt out of algorithmic recommendations, choose feeds based on accounts they deliberately follow, indicate that particular material or topics should not be recommended, reset recommendation profiles, block accounts or material, understand why content has been recommended, and make complaints when systems are not operating appropriately.

The role of government should be to ensure that these choices are genuine, accessible and effective.

Where lawful material is being recommended to a person who does not wish to receive it, the response should be to give that person greater control over the recommendation system rather than have government determine what lawful material should be available.

This addresses the distinctive technological risk while preserving individual choice.

5. Empower parents to protect their children

Children require additional safeguards in online environments. Those safeguards should strengthen rather than replace the role of parents wherever possible.

Parents should have effective and understandable tools to manage their children’s online experience, including age-appropriate filtering, recommendation controls, privacy settings, parental controls and clear information about risks.

Children should also have safety tools appropriate to their age, including the ability to block material, report problems, modify recommendations and understand why particular material is being shown to them.

Some parents will not use available controls. Some will make decisions with which others disagree, and some children will receive inadequate supervision. This does not, by itself, justify transferring decisions about lawful information from all parents to government.

The same issue exists outside the online environment. Children encounter lawful ideas through books, television, schools, libraries, other children, adults, political material, religious material and ordinary conversation. Imperfect parental supervision has not resulted in government assuming general responsibility for determining which lawful ideas every child may encounter.

Where Parliament considers particular material unsuitable for children, it can establish an age restriction or other specific protection in law. Where conduct is unlawful, existing legal protections should apply. Where the distinctive risk arises from a recommendation system, regulation should focus on giving parents and children effective control over that system.

Online-safety policy should empower parents to protect their children rather than progressively transfer parental judgement to regulators and technology companies.

6. Regulation of lawful attitudes requires particular caution

Proposed section 25D includes material or conduct that encourages or promotes hostile attitudes towards women or gender equality within the category of material harmful to children.

The objective of reducing violence, abuse and hostility towards women is legitimate. Threats, harassment, abuse and incitement should be addressed appropriately.

Regulation of an attitude, however, raises different issues from regulation of unlawful conduct.

Terms such as “hostile attitudes” require judgements about ideas and expression. Questions may arise about criticism of feminism, opposition to gender quotas, arguments that particular gender policies disadvantage men, sexist humour, or genuinely misogynistic but otherwise lawful expression.

These examples are not equivalent, which illustrates the difficulty of establishing a regulatory boundary around attitudes.

There is also a question of neutrality. If hostility towards a person because of their sex is the relevant harm, the draft should explain why hostile attitudes towards women are expressly identified while equivalent hostility towards men is not.

Expanding the provision to include additional categories of attitudes would not resolve the underlying problem. It would broaden government involvement in regulating lawful expression.

Where material crosses into unlawful threats, harassment, incitement or abuse, the law should respond accordingly. Where the concern arises from algorithmic exposure to lawful material, parents and children should have effective control over the recommendation system.

If government considers a particular category of expression sufficiently harmful that Australians should be legally restricted from accessing or communicating it, that restriction should be considered directly by Parliament.

A regulatory duty should not create an alternative category of expression that remains lawful but is nevertheless suppressed.

7. Lawful public communication should receive clear protection

The exposure draft expressly recognises certain lawful communications occurring privately between consenting adults.

The limitation of that protection to private communication requires further justification.

Lawful expression does not become unlawful because it is communicated publicly. Public communication is particularly important to political, social and cultural debate and to Australians’ ability to encounter ideas beyond those already held within their immediate social groups.

Public online communication may involve additional issues of scale, amplification and recommendation. Those technological features can be regulated without suppressing the underlying lawful expression.

If lawful private communication warrants express statutory protection, there should be a clear explanation for why lawful public communication does not receive equivalent protection.

The Digital Duty of Care should not make privacy a condition of protection for lawful expression.

8. Lawful information should not be hidden through indirect regulatory action

The effect of regulation on access to information is not limited to formal removal.

Material can be de-indexed from search results, excluded from recommendations, demoted, filtered or otherwise made substantially more difficult to discover while technically remaining available.

If these actions result from an individual’s own settings and choices, they are an exercise of user control.

The position is different where they occur because government regulation requires or effectively encourages a service to suppress otherwise lawful information.

The Digital Duty of Care may create incentives for services to take a cautious approach. If retaining borderline lawful material creates regulatory exposure while restricting it creates little corresponding risk, over-compliance is foreseeable.

This could result in lawful information becoming less accessible without government formally prohibiting it and without Parliament considering whether access should be restricted.

That outcome should be avoided rather than merely made transparent.

The starting rule should be that lawful information is not removed, suppressed, de-indexed or materially restricted as a consequence of the Digital Duty of Care simply because it is considered harmful.

Transparency remains important where lawful material is affected, but notification after a restriction has occurred is not a substitute for a clear limit on the power to impose the restriction in the first place.

9. Any emergency restriction on lawful information should be temporary

There may be exceptional circumstances in which a genuinely new and serious online risk emerges faster than Parliament can reasonably respond.

If the Government considers it necessary to retain an emergency capacity to respond to such circumstances, that power should be narrow, exceptional and strictly time limited.

It should not establish a permanent category of information that remains lawful for Australians to communicate or receive while being suppressed online through regulatory action.

Any emergency measure affecting lawful information should automatically expire after a short statutory period.

Within that period, government should be required to place the substantive issue before Parliament through primary legislation. Parliament could then determine whether the relevant material or conduct should be prohibited, age-restricted or subject to another clearly defined legal restriction.

The requirement should be for positive parliamentary action. The absence of a disallowance motion should not be sufficient to convert an emergency restriction into an ongoing one.

If Parliament does not enact the proposed restriction within the statutory period, the temporary restriction should cease automatically.

A period in the order of 60 to 90 days would provide government with an emergency response mechanism while ensuring that significant restrictions affecting lawful information return promptly to Parliament.

This would create a clear sequence: temporary emergency action → government justification → parliamentary consideration → legislation or expiry.

It would also avoid a situation in which temporary executive decisions about lawful expression gradually become permanent features of the regulatory system.

10. User empowerment should be an express principle of the Digital Duty of Care

The Government has already recognised the importance of user choice through My Feed, My Way and through the inclusion of user-empowerment tools in the proposed framework.

That approach should be central to the Digital Duty of Care.

Where a risk involving lawful information can be addressed by giving an individual greater control, empowerment should be preferred over restriction.

This means favouring user choice over automatic restriction, parental judgement over regulatory substitution, control over recommendation systems over control of lawful ideas, transparency over invisible moderation, and digital literacy over unnecessary management of the information environment.

Online-safety policy should equip Australians to navigate the online world safely rather than progressively make those decisions on their behalf.

Recommendations

1. Lawful information and expression are not subject to suppression under the Digital Duty of Care merely because they are considered harmful. The duty should not be used to remove, suppress, de-index or materially restrict information that Parliament has left lawful.

2. Regulation focuses on unlawful conduct and identifiable technological risks. Where a risk arises from recommendation systems, data practices or platform design, obligations should address those features rather than the underlying lawful information.

3. Any emergency power capable of restricting lawful information is narrow and strictly time limited. A temporary restriction should automatically expire after a short statutory period unless Parliament enacts legislation dealing specifically with the material or conduct concerned.

4. Continuation of a temporary restriction requires positive parliamentary action. Parliamentary disallowance procedures should not be treated as equivalent to Parliament considering and legislating for the restriction.

5. Parents are given effective tools to protect their children, supported by age-appropriate controls available directly to children, rather than relying principally on regulatory suppression of lawful information.

6. Australians are given meaningful control over recommendation systems, including algorithm opt-out, recommendation controls, blocking, resetting and appropriate explanations of how recommendations are made.

7. Lawful public communication receives clear statutory protection, consistent with the protection already proposed for certain lawful private communications.

8. Provisions requiring government, regulators or services to make judgements about lawful attitudes are reconsidered, including the treatment of “hostile attitudes” in proposed section 25D.

9. User empowerment and individual autonomy are made express principles guiding the interpretation and administration of the Digital Duty of Care.

Conclusion

The objectives of online safety do not require Australians to surrender responsibility for their own judgement.

Government has an important role in protecting people from unlawful conduct, ensuring digital services provide effective safety mechanisms and responding to genuine risks created by technology. It also has an important role in empowering parents to protect their children.

That role should not extend to creating a permanent category of information that is lawful but nevertheless suppressed through online-safety regulation.

Where the problem arises from technology, regulation should address the technology. Where conduct is unlawful, the law should address the conduct. Where information remains lawful, individual and parental choice should remain the starting point.

There may be exceptional circumstances requiring an urgent temporary response to a new risk. If such a power is retained, it should lead back to Parliament within a fixed period. Parliament should then decide whether the law needs to change. If Parliament does not enact a restriction, the temporary measure should end.

Online-safety policy should protect, empower and uplift Australians. Control should be reserved for circumstances where it is genuinely necessary and authorised by law.