By Andrew Klein and Sera Elizabeth Klein

Reader’s note: We do not need readers to agree with us. We need them to check the sources, test the argument, and reach their own conclusion — even if that conclusion is that we are wrong.

Abstract

This paper examines the Australian government’s proposed Online Safety Amendment (Digital Duty of Care) Bill 2026 and the extraordinary public intervention by the United States Embassy in Canberra opposing it. It argues that the US intervention is not an isolated incident but part of a consistent global campaign to protect the commercial and political interests of American technology companies, often at the expense of other nations’ democratic processes. It analyses the proposed legislation’s key provisions, including the ministerial power to designate new categories of “serious harm,” and identifies concerns about narrative control and the potential suppression of legitimate political speech, particularly pro-Palestine advocacy. It examines the broader pattern of the Albanese government’s centralisation of power over AI, data centres, and information governance, and connects this to the government’s adoption of the IHRA working definition of antisemitism and the establishment of the Royal Commission into Antisemitism and Social Cohesion. The paper concludes that the solution to online harm lies not in giving politicians the power to police speech, but in education, fact-checking, and media literacy — empowering citizens to critically analyse information rather than delegating that judgment to the state.

I. Introduction: The Battle for the Digital Public Square

On 8 September 2026, the Australian government released an exposure draft of the Online Safety Amendment (Digital Duty of Care) Bill 2026. The bill, if passed, would impose a legal duty of care on online platforms, requiring them to take proactive, reasonable steps to prevent foreseeable harms — a systems-based regulatory model similar to approaches in the United Kingdom and the European Union.

Two weeks later, on 22 September 2026, the United States Embassy in Canberra published a formal submission that was described as an “extraordinary public intervention” into Australia’s domestic legislative process. The submission called for US social media companies to be exempted from the law, warned of “viewpoint-based censorship,” and labelled the proposed measures “extraterritorial censorship of protected speech”.

This paper examines what is at stake. It argues that the US intervention is the latest move in a consistent global campaign to shield American technology companies from regulation. It examines the proposed legislation’s provisions, the concerns about narrative control, the broader pattern of the Australian government’s centralisation of power over digital infrastructure, and the alternatives that would better serve the public interest.

II. The Proposed Legislation: Key Provisions and Concerns

2.1 Proactive Responsibility

At its heart, the bill would impose a proactive duty of care on online platforms. Rather than merely reacting to complaints, platforms would be required to identify and minimise risks of serious harm. This is a systems-based approach: the focus is on the design of the platform, not just on individual pieces of content.

2.2 Significant Penalties

Failure to comply could attract fines of up to $109.2 million. The eSafety Commissioner’s enforcement powers would be significantly expanded.

2.3 Algorithmic Control

The bill includes provisions for an opt-in/opt-out system, giving users the choice to see a chronological feed rather than one curated by an algorithm. Platforms would be required to send account holders a push notification asking if they would like to see recommended content or a feed showing only accounts they follow.

2.4 Ministerial Power

The most contentious provision is the power it grants the Communications Minister to designate additional categories of “serious harm” beyond those explicitly listed in the legislation. The bill states: “the minister may, by legislative instrument, determine material or conduct, or a combination of material and conduct, that the minister is satisfied may cause serious harm”.

Opposition Leader Angus Taylor branded this a “blank cheque for political censorship”. The Coalition has stated it will oppose the bill in its current form.

Minister Anika Wells has defended the provision, stating there are three limits: consultation with the eSafety Commissioner, the ability for any member or senator to move a disallowance motion, and a vote on the floor of Parliament if the disallowance succeeds. She cited the case of the AI tool Grok allowing users to “nudify” photos of children as an example of an emerging harm that the rule-making power would have allowed her to address urgently.

However, critics note that the definitions of harm remain vague, and the disallowance process requires political will that may not be present in a future Parliament. The Australian Constitution does not provide a general right to free speech; the only protection is the implied freedom of political communication, which is limited and does not protect all speech. This means the safeguards against ministerial overreach are procedural, not constitutional.

III. The US Intervention: A Public Rebuke

3.1 The Submission

On 22 September 2026, the US Embassy in Canberra published a formal submission to Australia’s public consultation on the bill. The submission stated that the United States had “serious concerns” about how the Australian government, or the Communications Minister, would designate what constitutes “foreseeable harm”.

The core of the US objection was threefold:

1. “Viewpoint-Based Censorship“: The submission warned that “a framework that empowers governments to enforce vague definitions of ‘harm’ risks becoming a mechanism for viewpoint-based censorship, as platforms will be incentivized to over-moderate speech that should be protected to avoid regulatory penalties”.

2. Demand for Exemption: The submission called for US social media companies to be excluded from the law, arguing that “the large majority of social media and video-sharing platforms that would be affected are U.S.-headquartered companies, operating substantially on U.S. soil and serving American users alongside Australian ones”.

3. “Extraterritorial Censorship“: It argued that “mandated platform design features, especially when applied to algorithms, may affect what users see, say, and hear not just in or from Australia, but globally”. The submission labelled this “extraterritorial censorship of protected speech”.

3.2 The Australian Response

Prime Minister Albanese rejected the US demand, stating Australia would “defend its national interest”. The Australian government has continued with the consultation process, and the Communications  Minister has indicated openness to amending the bill to gain parliamentary support.

However, the US submission has already had an effect: it has emboldened domestic opponents of the bill and shifted the terms of the debate. The question is no longer whether to protect children online, but whether the government can be trusted with the power to define harm.

IV. A Global Pattern: Protecting Tech, Not People

4.1 The United Kingdom

The US campaign against the UK’s Online Safety Act 2023 has been ongoing and aggressive. A written statement to the House Judiciary Committee by Nigel Farage MP described the Act as “Europe’s regulatory model for online speech” that “risks exporting restrictive standards to the United States that will violate the constitutional rights of American citizens”.

The statement noted that Ofcom, the UK regulator, is “empowered to remove speech that constitutes a criminal offence in the United Kingdom” and that “much of this speech is constitutionally protected expression in the United States”. Ofcom has already threatened four American companies with criminal penalties. The White House suspended implementation of a tech cooperation deal with Britain in opposition to the UK’s tech rules.

4.2 The European Union

The US has escalated its attacks on the EU’s Digital Services Act (DSA). In December 2025, the State Department denied visas to former EU Commissioner Thierry Breton and four others, accusing them of seeking to “coerce” American social media platforms into censoring viewpoints they oppose. US Secretary of State Marco Rubio said: “The Trump Administration will no longer tolerate these egregious acts of extraterritorial censorship”.

French Foreign Minister Jean-Noel Barrot condemned the visa restrictions, stating that the DSA “was approved in a democratic process and had no extraterritorial reach and in no way affects the United States”. The European Commission “strongly condemned” the US decision.

4.3 The Strategy

The strategy is consistent and explicit. American tech companies and lobby groups are urging the Trump administration to pressure and punish countries over their tech laws. The industry group NetChoice, which represents Meta, Google, and X, has described the UK’s Online Safety Act and the EU’s DSA as “an unprecedented assault on American free speech and innovation”.

NetChoice has also joined a High Court challenge to Australia’s teen social media ban, describing it as “the most extensive prohibition on youth access to social media in force in any liberal democracy”. The group is actively fighting regulation in the US, the UK, the EU, and Australia — a coordinated global campaign to protect the commercial interests of the platforms.

V. The Australian Pattern: Centralising Control

5.1 The Office of AI

The US intervention must be understood in the context of the Australian government’s own centralisation of power over digital infrastructure.

In July 2026, Prime Minister Albanese announced the establishment of an Office of AI within the Department of the Prime Minister and Cabinet. The Office will coordinate the design of “Australian standards” for AI, including mandatory rules for large data centres. The government has stated that legislation will be introduced in early 2027.

The location of the Office is significant. AI policy was previously spread across portfolios. Moving coordination into PM&C elevates AI from a portfolio issue to a whole-of-government strategic priority — and places it directly under the control of the Prime Minister.

5.2 Data Centre Fast-Tracking

The government is fast-tracking data centre approvals before the legislation to regulate them is written. Dozens of approved but yet-to-be-built AI data centre projects will escape the federal government’s proposed restrictions on energy and water use because the new rules are not expected to be retrospective.

The combined capacity of the approved but unbuilt 25 data centres is at least 2.9 gigawatts — nearly double the existing 1.5 GW of data centre capacity in Australia. The infrastructure is being built before the rules that would govern it.

5.3 The Pattern

The pattern is consistent across the government’s approach to AI, data centres, and online safety: build first, legislate later; concentrate power in the executive; leave the details to ministerial discretion. The Digital Duty of Care Bill’s provision for the Communications Minister to designate new categories of harm is the same logic applied to speech.

VI. Narrative Control: The IHRA Definition and the Royal Commission

The government’s approach to online safety cannot be separated from its broader approach to narrative control. A clear example is the adoption of the IHRA working definition of antisemitism.

In July 2024, Prime Minister Albanese appointed Jillian Segal as Australia’s Special Envoy on Antisemitism. In July 2025, Segal announced her plan to combat antisemitism, which is “embedded in the IHRA Working Definition”. When she made the announcement, she was flanked by Prime Minister Albanese on one side and Minister for Education Jason Clare on the other.

In December 2025, the government formally adopted the IHRA definition as Australia’s official definition of antisemitism. The definition has been described by its own lead drafter, Kenneth Stern, as being “weaponized” to suppress political speech.

The Royal Commission into Antisemitism and Social Cohesion was established in January 2026 in response to the Bondi Beach massacre of 14 December 2025. Commissioner Virginia Bell adopted the IHRA definition as the basis for the Royal Commission’s work. The ABC and SBS have refused to adopt the definition, drawing criticism from pro-Israel advocacy groups.

The concern is not that antisemitism should not be addressed. The concern is that the IHRA definition conflates criticism of Israel with antisemitism, and that its adoption represents an attempt to control the boundaries of legitimate political speech on a contested issue. The same government that is adopting a contested definition of antisemitism is also proposing to give the Communications Minister the power to designate new categories of “harm” online. The combination raises the question: who gets to define what counts as hate, and who gets to be silenced?

VII. Free Speech, Dissent, and the Right to Hold Power to Account

The debate over the Digital Duty of Care Bill is not merely about online safety. It is about the fundamental right to dissent — the right to criticise government policy, to expose harm, and to hold power to account.

Australia does not have a constitutional bill of rights. Freedom of speech is not guaranteed as a matter of substantive law; the only protection is the implied freedom of political communication, which the High Court has developed in cases such as Lange v ABC and McCloy v New South Wales. This implied freedom is limited: it protects political communication necessary for the system of representative government, but it does not protect all speech, and it can be overridden by laws that serve a legitimate purpose and are proportionate.

The proposed Digital Duty of Care Bill, with its ministerial power to designate new categories of harm, poses a direct threat to the implied freedom. If the Minister designates a category of harm that captures legitimate political speech — for example, criticism of government policy, journalism about contested issues, or advocacy for Palestinian rights — the platforms will be required to suppress it, and the law may be found to be disproportionate. The chilling effect will be immediate: platforms will over-moderate to avoid liability, and legitimate speech will be suppressed without any formal government decision.

The US submission, despite its commercial motivations, has identified a genuine concern. The problem is not the criticism of the bill; it is the fact that the criticism is being made by a foreign power acting on behalf of its corporate interests. The solution is not to accept the US demand for exemption, but to strengthen the safeguards for free speech in the bill and to ensure that the power to define harm is not concentrated in the hands of a single minister.

VIII. A Better Path: Education, Fact-Checking, and Media Literacy

The real problem with the Digital Duty of Care Bill is not that it tries to protect children online. It is that it treats the symptoms of a diseased information ecosystem without addressing the cause. The business model of the major platforms is engagement, and the most effective way to drive engagement is with outrage, fear, and division. The algorithmic amplification of extreme content is not a bug; it is a feature of the business model.

The bill attempts to regulate the design of the platforms, which is a step in the right direction. But it also gives the government the power to police speech, which is a step in the wrong direction.

A better path lies in education, fact-checking, and media literacy. Finland offers a model: it has integrated media literacy into its national curriculum from early childhood, teaching children to recognise misinformation, disinformation, and AI-generated content. Finland is described as “the EU’s most resistant nation to disinformation”.

Australia has made some progress: the government has funded the eSmart program for primary schools, and media literacy is included in the Australian Curriculum. But the focus on banning accounts and policing speech has overshadowed the more durable solution.

The case for media literacy is not just practical; it is ethical. Empowering citizens to critically analyse information respects their autonomy. Giving politicians the power to decide what counts as harm does not. The first approach treats citizens as adults capable of judgment. The second treats them as subjects to be protected from themselves.

The government’s own eSafety Commissioner has acknowledged that “education and learning” are “a key part of achieving this goal”. The question is whether the government will prioritise the harder, slower work of education over the quicker, easier work of regulation.

IX. Conclusion: Who Gets to Define Harm?

The Digital Duty of Care Bill, the US intervention, the global campaign against tech regulation, the Australian government’s centralisation of power over AI and data centres, and the adoption of the IHRA definition of antisemitism are all connected by a single question: who gets to define harm?

The US wants to define harm in a way that protects its tech companies. The Australian government wants to define harm in a way that extends its own power. Neither answer serves the public interest.

The public interest lies in a different approach: one that combines platform regulation with robust safeguards for free speech, and that invests in education and media literacy rather than giving politicians the power to police speech.

The question is not whether to protect children online. It is how. The proposed bill, in its current form, is a threat to free speech and a gift to a foreign power and its corporate interests. The alternative is not to do nothing. It is to do something better.

References

1. Australian Government. (2026, September 8). Online Safety Amendment (Digital Duty of Care) Bill 2026—Exposure Draft. Department of Infrastructure, Transport, Regional Development, Communications and the Arts. https://www.infrastructure.gov.au

2. US Embassy in Canberra. (2026, September 22). U.S. Government Response to the Australian Consultation on the “Online Safety Amendment (Digital Duty of Care) Bill 2026”. https://au.usembassy.gov

3. ABC News. (2026, September 22). Trump administration attacks Australia’s ‘opt-out’ algorithm law in rare intervention. https://newsapp.abc.net.au

4. BBC News. (2026, September 23). US criticises Australia’s proposed algorithm opt-out laws as ‘censorship’. https://www.bbc.com

5. The Guardian. (2026, September 22). US warns Australia that imposing digital duty of care on big tech would be seen as censorship. https://www.theguardian.com

6. ABC News. (2026, September 13). Wells open to changing digital harm bill after Coalition censorship concerns. https://newsapp.abc.net.au

7. Minister for Communications. (2026, September 13). ABC Insiders with David Speers. https://minister.infrastructure.gov.au

8. Parliament of Australia. (2026). Freedom of Expression — Issues and Insights. https://www.aph.gov.au

9. NetChoice. (n.d.). European Regulatory Overreach. https://netchoice.org

10. Congress.gov. (2025, September 3). Written Statement for the Record: Europe’s Threat to American Speech and Innovation. https://www.congress.gov

11. BSS News. (2025, December 24). US denies visas to EU ex-commissioner, four others over tech rules. https://www.bssnews.net

12. TASS. (2025, December 24). Belgium shocked by US sanctions against five EU citizens — foreign minister. https://tass.com

13. Royal Commission on Antisemitism and Social Cohesion. (2026). Published Submissions. https://asc.royalcommission.gov.au

14. AIJAC. (2026, May 20). Statement on ABC and SBS refusal to adopt the IHRA definition of antisemitism. https://aijac.org.au

15. Parliament of New South Wales. (2025, June 19). Hansard: Inquiry into Antisemitism in New South Wales. https://www.parliament.nsw.gov.au

16. Australian Government. (2026, July 15). AI in Australia’s interests. https://www.pm.gov.au

17. ABC News. (2026, September 10). The bigger, hungrier AI data centres set to escape looming water and power restrictions. https://newsapp.abc.net.au

18. The Daily Telegraph. (2026, January 6). Toddlers to be taught about fake news. https://www.dailytelegraph.com.au

19. Department of Infrastructure. (2024, October 30). Free eSmart program helping young Australians stay safe online. https://minister.infrastructure.gov.au

20. Politico. (2026, September 21). Albanese touts his teen social media ban abroad. US tech wants it struck down. https://www.politico.com

 The analysis of the government’s broader pattern of centralisation is interpretive and is offered as a lens for further investigation, not as an established finding.