Foreword
In the day that the 2024 parliamentary session concluded, more than 30 bills were passed, as the government delivered on its ‘clear message’ to get important legislation through. Two bills that stand out for their profound impact on free speech are those addressing social media regulation and privacy. A Misinformation bill was abandoned at a late stage, to prevent it being a spanner in the works of that historic accomplishment. At the beginning of the next parliamentary session, in February 2025, a criminal hate speech bill passed that was similarly far-reaching. This article considers that legislation, and related executive action concerned with online safety, in light of the cost to the freedoms that support the rule of law and lie at the heart of a democratic society. It assesses how rule of law principles play their part in highlighting excessive power, and distinguishing good from bad law, including by the yardstick of basic human rights norms. The tools of accountability that support the rule of law are enhanced when human rights are properly protected, especially freedom of expression and related rights. When they are undermined, so are crucial aspects of the rule of law.
In Australia, opinions differ on the scope of the rule of law. Some consider it to be confined to Lord Bingham’s encapsulation that “all persons and authorities within the state, whether public or private, should be bound by and entitled to the benefit of laws publicly and prospectively promulgated and publicly administered in the courts.”[1] However, it does not stand in isolation, since he extruded eight principles from it. These are:
- the law must be accessible, intelligible, clear and predictable;
- rights and liabilities should be resolved by application of the law and not the exercise of discretion;
- laws should apply equally to everyone, except where objective differences justify differentiation;
- officials must exercise their powers conferred in good faith, fairly, for the purpose for which the powers were granted, within power, and not unreasonably;
- the law must afford adequate protection of fundamental human rights;
- civil disputes should be resolved without inordinate cost or delay;
- there must be procedural fairness in the application of the law; and
- the rule of law requires compliance by the State with its obligations in international law as in national law.[2]
These principles are generally widely acknowledged, except possibly the last, given Australia is a dualist nation state, and some are comforted by the fact that treaties have little or no effect until implementing legislation has been enacted. This ensures that Parliament remains the ultimate legislative authority. Yet Lord Bingham also made the point that “[t]he means by which an obligation becomes binding on a State in international law seem to be quite as worthy of respect as a measure approved, perhaps in haste and without adequate enquiry, perhaps on a narrowly divided vote, by a national legislature.”[3]
It is therefore fair for me to assume for the purposes of this paper that the rule of law requires compliance by the State with its obligations in international law, including those related to human rights, especially since certain treaty rights — notably the freedom of expression and related rights — play a vital function in support of the rule of law.
Introduction
At Ivy League sporting events, Yale – which rejoices in the motto lux et veritas (light and truth) — indulges in a terrace chant at the expense of Harvard — which can only boast veritas. It goes: “Your veritas sucks if you ain’t got lux”.
A similar taunt directed at government could be rendered, “Your lex sucks if it ain’t got ius”. This is because there is a crucial distinction, familiar to legal philosophers, between the Latin lex, meaning ‘law’ in the form of legislation (even if it is bad law), and ius, broadly translated as ‘right’, encompassing overreaching principles of justice and fairness. It enables legislation to be judged as good or bad, not just whether it is within power.[4]
At its simplest, the rule of law holds that no one is above the law. Viewed through a political prism, the prominent New Zealand legal philosopher Jeremy Waldron considered its most important demand to be that
“people in positions of authority should exercise their power within a constraining framework of well-established public norms rather than in an arbitrary, ad hoc, or purely discretionary manner on the basis of their own preferences or ideology”.[5]
This article considers broader aspects of the rule of law, including those which establish certain requisite properties of law. The ‘rule of law’ is a principle with a number of elements which help identify shortcomings in the exercise of power, or in the law itself, and requires accountability to the law by everyone and equal application of the law to everyone. In this way we can better gauge when government enacts legislation or takes executive action which is not ‘just or fair’, measured against certain expectations of a democratic society, including those set by established human rights standards.
Rule of law principles generally operate at a high level of abstraction but with concrete support in law for certain facets. For example, in Australia the Commonwealth Constitution achieves the familiar separation of the legislature, executive and judiciary, designed to avoid the concentration of power, and thereby abuse, within any one branch of government. Administrative law establishes avenues for reviewing the exercise of executive power, to ensure government decision-making remains within power, and is transparent and accountable. The common-law ‘principle of legality’ ensures that ambiguity in a statute is interpreted in favour of fundamental common law rights and freedoms, where possible, by expecting parliament to use clear and unambiguous language when restricting those freedoms.
However, other crucial aspects of the rule of law are not as well instantiated in Australian law. One especially important element of the rule of law concerns the compatibility of legislation with international human rights norms and standards. As the UN Secretary-General Kofi Annan explained in 2004:
“The ‘rule of law’ is a concept at the very heart of the Organization’s mission. It refers to a principle of governance in which all persons, institutions and entities, public and private, including the State itself, are accountable to laws that are publicly promulgated, equally enforced and independently adjudicated, and which are consistent with international human rights norms and standards.”[6]
In a 2023 New Vision for the Rule of Law, the current UN Secretary-General António Guterres emphasised that the rule of law calls for “unwavering respect for all human rights … Human rights and the rule of law are mutually reinforcing — the advancement of the rule of law is essential for the protection of all human rights, and human rights are central to the rule of law”.[7]
The understanding of the rule of law in Australia, as recounted in the Law Council of Australia’s policy statement on the rule of law, takes specific account of the need for consistency with international human rights standards, stating that international legal obligations must be met, whether created by treaty or arising under customary international law. It also particularises other long-established rule of law stipulations established by Lord Bingham. .[8] This raises a high expectation of conformity of legislation with fundamental human rights, of transparency and accountability, and fairness in the application of the law, in addition to the structural separation of powers and administrative review procedures already mentioned.
This article considers areas of possible ‘drift’ from rule of law principles in recent times, particularly concerning fundamental human rights, in an age when rights are susceptible to being put into political service, rather than to promote the comprehensive enjoyment of all rights by all Australians on equal terms. It takes for the purpose of illustration an assortment of recent enactments, challenges to the use of executive authority, scrutiny and other democratic and review processes, and inquiries. It examines whether there are any notable anomalies which, applying established rule of law principles, might signal the need for correction. Put another way, it asks whether we have reached the point at which the rule of law, and the protection for fundamental human rights as an element supporting it, if not also integral to it, need to be reclaimed.
Some basic human rights principles
The different perspectives of the political parties at any time in power produce different priorities and legislative outcomes, needless to say. Views inevitably differ on how power should be exercised to achieve societal well-being, or as Aristotle put it, to “secure the good life”.[9] However, a common standard applies to the protection of human rights, established by treaty, which is the duty of every government, regardless of its political complexion, to uphold for everyone under Australian jurisdiction. This is an abiding principle of responsible government aside from rule of law principles.
In Australia, human rights are protected through a combination of legislation, including Constitutional provisions, and evolving common law. Some legislation is specifically enacted to give effect to rights in international treaties, such as those dealing with discrimination on various grounds, and the protection of children. International human rights obligations often play a vital role in providing the ostensible justification for legislation, including where it imposes quite serious restrictions on rights. It is all the more important that the rights analysis which is undertaken in explanatory memoranda, statements of human rights compatibility and scrutiny reports fulfil their task in defending rights against excessive intrusion, or at least exposing where this is occurring. Rights have become embedded in Australian lawmaking practice in this sense, as well as in substantive law for most rights.
There is a regular throughput of legislation needed to meet a societal challenge or emerging risk in a way that shores up protection for fundamental rights that may be threatened. This often involves a number of rights-holders, and when the legislation puts their interests in tension, their ‘competing rights’ are required to be ‘balanced’.
When done conscientiously, the legislative art is to settle on a scheme which among other things blends a number of essential requirements. These are to ensure: that the full scope is recognised of all rights engaged; that no right is restricted beyond the scope allowed for it; and that all rights in play are ‘maximised’. These are not merely aspirations but are meant to direct outcomes. As explained by Heiner Bielefeldt, the task involves preserving the substance of human rights, to the maximum degree possible.[10] Bielefeldt is not keen on ‘balancing’ language (nor is this author) since it conceals the fact that the burden of justifying any restrictions on a fundamental right remains firmly on the government. He also stresses that all too often the task of maximising rights is abandoned too easily.
In practice all of these ends should be pursued in the formulation of legislation, in judicial decision-making, and in the exercise of statutory powers, whenever different rights are engaged. This is because all rights in the International Covenant on Civil and Political Rights (ICCPR) are meant to be given full effect.[11] No single right may be favoured over another, or expanded beyond its definitional boundaries in a way that undermines another right. The definition of each right, and what is permissible by way of restriction, are determined by the treaty in which the relevant right is found. For example, ‘freedom of expression’ is defined in article 19 of the ICCPR which specifies that freedom of opinion cannot be limited at all, but that the expression of opinion may be, where necessary, on particular grounds (such as the rights of others, national security, public order, and public health). Privacy is protected by article 17 in a somewhat different way, against ‘arbitrary interference’. Each right is cast in different terms.
Although most ICCPR rights are fully respected in Australian law, there are exceptions, in spite of 45 years having elapsed since Australia ratified the ICCPR. A prominent example is freedom of expression, which cannot meaningfully be asserted on the basis of substantive law to resist restrictions which overstep the ICCPR limits just mentioned. Instead, there is at best only ‘free speech’ protection, made up of two main elements. First, there is the implied freedom of political communication under the Commonwealth Constitution,[12] which is not a personal right but a much narrower brake on the power of government to regulate political expression (and even that has recently been put in serious doubt).[13] The second component of ‘free speech’ is the freedom that generally exists in the unregulated space to do what is not specifically prohibited by legislation.[14] In other words it exists only until encroached upon by legislation. In this article, ‘free speech’ is the term used to refer to this limited form of protection in Australia (with these two elements), while ‘freedom of expression’ is used to refer to the standard that should apply in Australia under the ICCPR, but does not yet, in terms of substantive law.
The disparity between free speech and freedom of expression is important to observe, especially since it is a fundamental obligation of responsible government, and of Australia as a party to the ICCPR, not to encroach upon the treaty standard of ‘freedom of expression’. In the absence of proper protection for that freedom in Australia, it is more incumbent on government to support and not undermine it when fulfilling its political mandate.
The freedom of expression also has a special role in giving practical effect to rule of law principles, especially in its capacity to hold to account the misuse of authority, to expose corruption and lies by officials, and is integral to a functional democracy; but also because unjustifiable restriction of the freedom signals a rule of law failure, given that the “rule of law calls for unwavering respect for all human rights”, as already noted.
The following discussion illustrates rule of law principles as they relate to recent proposals for misinformation legislation, court challenges to the exercise of power by the e-Safety Commissioner, legislation protecting privacy and against hate crimes, parliamentary process and certain inquiries.
Rule of law principles as they relate to recent government initiatives
Misinformation regulation
In 2023 and 2024 the government proposed two different bills to address the ‘serious harm’ caused or contributed by ‘misinformation’ and ‘disinformation’ affecting (among other things) the electoral process, public health, emergency services and the Australian economy. Both would have imposed very significant restrictions on free speech. The first bill received such stinging public opposition from certain quarters, that when all went quiet for many months it may have been assumed the whole idea had been dropped. It was revived in a new bill which was presented to Parliament in the second half of 2024, and was only abandoned once it became clear it had ‘no pathway’ in the Senate.[15]
Both bills were extremely complex, and would have incentivised social media and other digital platforms to over-censor content to avoid eye-watering fines. Unlike comparable legislation overseas, it crucially failed to apply any meaningful support for freedom of expression (as understood by the ICCPR) or even to preserve the much lower standard of protection for free speech in Australia. (EU legislation, for all its shortcomings, at least nominally preserved the European Convention’s ‘freedom of expression’).
Although the bills had the ostensible aim of protecting Australians against harm, one reason they were opposed was because they were themselves seen to be capable of material harm. By largely silent and unseen mechanisms the legislation would have meant that perfectly harmless information and opinions, as well as online content that was highly valuable in correcting a harmful official narrative, or holding to account those in power or guilty of corruption, could in practice be excluded from view by applying the tag ‘misinformation’ or ‘disinformation’. Among the content likely to be excluded would be challenges to controversial policies, practices, evidence and suppositions. A commonplace purpose would be to contradict, or expose the shortcomings of any incumbent narrative. The bills would have tended to produce a less enlightened population, more likely perhaps to yield to official versions of the truth that were not so visibly contested. Views that demolish false narratives would not see the light of day.
Misinformation regulation was also likely to increase the risk of political bias in censorship. (The likelihood of such a risk caused the UK’s Online Safety legislation to include specific provisions to address political bias.) The ‘Australian Twitter Files’ scandal had already demonstrated how digital platforms had been working with the Australian Department of Home Affairs to remove social media content on thousands of occasions, even though the posts in question were not harmful in nature, and their content apparently turned out to be true.[16] One post was a joke that went to an issue in open debate at the time, and depicted then Victorian premier Daniel Andrews in a mask bearing the inscription, “This mask is as useless as me”.[17]
Although the proposed legislation was abandoned, digital service providers continue to be regulated by the Online Safety Act 2021, and are charged through industry codes and standards set by the eSafety Commissioner with combating ‘misinformation’ and ‘disinformation’. The eSafety Commissioner’s powers concerning the regulation of online content are considerable, and were enhanced around the time that the misinformation bills were proposed (raising questions as to the necessity of those bills).[18] The scope and use of the eSafety Commissioner’s powers give rise to several concerns in the nature of transparency and accountability, as well as the certainty and predictability of the law.
Social media minimum age legislation
The Online Safety Amendment (Social Media Minimum Age) Bill was passed at the end of November 2024 in the chaotic circumstances described below (under the heading The proper functions of Parliament), in which a vast number of bills were processed concurrently in the last remaining sitting days before Christmas. Where legislation is proposed by government in excess of constitutional authority it does not automatically indicate a rule of law issue, but it is more likely when combined with other significant shortcomings which compromise transparency and accountability.
There are various ways in which the government used the best interest of the child to justify the social media minimum age legislation, invoking numerous ICCPR and Convention on the Rights of the Child rights which were said to be threatened. Among the serious harms the legislation claimed to be protecting against was the risk of children being exposed to false and misleading information that is prolific on social media.[19] It therefore adds further restriction to the pre-existing censorship of content on grounds of ‘false and misleading information’ by social media platforms under the Online Safety Act.
There may be constitutional law grounds for challenging the 2024 changes, for infringing on the implied freedom of political communication, which has already been mentioned as a component of free speech in Australia. The implied freedom allows Australians “to exercise a free and informed choice as electors”.[20]
The Explanatory Memorandum described the legislation as “a strong regulatory response” to safeguarding “the health and wellbeing of children”.[21] Yet as Prof Anne Twomey has pointed out, it does virtually nothing effective to prevent children from harm.[22] There is a complete mismatch between protection against harm claimed for the measure, and the legal effect it actually has. The government identified the harm to young people in the Explanatory Memorandum,[23] and more powerfully in its public campaigning, as including excessive screen time, social isolation, sleep interference, poor mental and physical health, and low life satisfaction. However, all it does is establish a minimum age for social media use, and require providers of an age-restricted social media platform to take reasonable steps to prevent age-restricted users from having an account with them. Such a clear non-alignment of the declared aim and operation of legislation invites speculation as to possible undeclared purposes it may serve.
The impact of the legislation on children in accessing quite harmless content is significant, and it includes not being exposed to a wide spectrum of opinions and perspectives, which is more important when a predominant influence in shaping their minds is public education, provided within the bounds of the National Curriculum and other educational standards.
One aspect of the test which determines whether there is an unjustifiable burden imposed on the implied freedom is whether it is “reasonably appropriate and adapted to serve a legitimate aim and in a manner that is compatible with the maintenance of the constitutionally prescribed system of representative and responsible government”. It requires the law to have a rational connection to its stated legitimate purpose, and for there not to be any obvious and compelling alternative measures that could achieve the purpose with a less restrictive effect. Also, there must be an appropriate balance between the purpose and the restriction imposed. The vulnerability to a successful constitutional claim is, according to Prof Twomey, that this measure does not appear to be suitable, necessary and proportionate, as required, given the obvious gap between the objectives claimed and the operational effect of the law.
That legislation was the responsibility of government, but an additional cause of growing concern is the work of the eSafety Commissioner.
The powers of the eSafety Commissioner
The office of the eSafety Commissioner was created by legislation in 2015 to protect children against targeted cyber bullying.[24] In 2017, the role was enlarged to cover online safety for all Australians, not just children. In 2021 the legislation was reviewed, and the Online Safety Act 2021 was put in place to respond to known online harms such as adult cyber-abuse, non-consensual sharing of intimate images, and seriously harmful online content. It is under the auspices of this expanded legislation that the eSafety Commissioner now exercises considerable power.
Unofficial signs that the office of the eSafety Commissioner may not be aligned with appropriate human rights standards include the eSafety Commissioner’s controversial comments at a World Economic Forum (WEF) event in Davos in 2022, that, “I think we’re going to have to think about a recalibration of a whole range of human rights that are playing out online”, including freedom of speech.[25]
Selective content suppression?
X Corp’s challenge to the eSafety Commissioner’s removal of footage of the stabbing of Bishop Mar Mari Emmanuel (while conducting a service at the Christ the Good Shepherd Church in Sydney) revealed how far the eSafety Commissioner was prepared to go — in pressing for a global takedown.[26] The Federal Court found it sufficient that X Corp only ‘geo-blocked’ the footage. The official claim that removing the footage was to avoid potential harm lacked force as equally or more graphic depictions of violence were left untouched, such as of the fatal stabbing at Bondi Junction around the same time. Also, there were available ways to preserve rather than delete the entire account of the bishop’s stabbing in the footage, for example by excising only the depiction of violence. Blocking the footage effectively removed an important news story, one angle of it being that the intolerance for certain religious beliefs has reached the point that it attracts such violence.
Action in excess of legal authority
The proceedings brought by Celine Baumgarten over removal of a tweet on a ‘queer club’ in a primary school exposed how the eSafety Commissioner had side-stepped formal statutory procedures for a ‘removal notice’ (by which a platform is ordered to remove content). The eSafety Commissioner acted in the mistaken belief that by issuing an informal ‘alert’ instead, the higher scrutiny and review jurisdiction of the Administrative Review Tribunal would be avoided.[27] It transpires that only a small handful of formal removal notices were issued in a year, but hundreds of informal ‘alerts’. This has been described as an attempt by the eSafety Commissioner to avoid the legal requirements of the Online Safety Act by taking action ‘outside’ it.[28] It is a fundamental rule of law principle that those entrusted with power are subject to the law and must operate within the boundaries of their authority.
Extension of age-related regulation to YouTube on the basis of ‘research’
Public trust in government and its institutions is vital to the successful implementation of good law. It is undermined when the public is misled by officials, and even when official statements give the appearance of artifice.
The social media requirement for under 16s, discussed above, was controversial because its effectiveness was in doubt, and the age verification processes were seen by some as presaging the normalisation of digital ID for everyday access to facilities, as a serious privacy issue. It was already a sensitive topic when in 2025 the eSafety Commissioner urged the government to extend the social media minimum age legislation to YouTube “because of the evidence of harm happening”. The logic which connects YouTube to social media platforms was “the known risks of harm on YouTube [and] the similarity of its functionality to the other online services”. In a National Press Club address at the time the eSafety Commissioner explained that YouTube had mastered “persuasive design features” which drive users down algorithmic rabbit holes.[29] However, a subsequent freedom of information request revealed that the research in question consisted of self-reported survey data (rather than independent observation or assessment) and the resulting 23-page report mentioned nothing at all about ‘rabbit holes’ or algorithms. On any reckoning it is arguably not ‘evidence’ fit to justify such far-reaching limits on access to information. It has been described as “massive overreach” which “distracts from genuine online safety efforts”.[30] It turns out the incidence of harassment, bullying and other harmful behaviours on YouTube was only 3% across all metrics, a tiny fraction of the numbers reported on social media platforms (e.g. 23% [for grooming] and 35% [for sexual harassment] on Snapchat).[31] It remains to be seen how this will play out, but the incident dented public trust among some in the eSafety Commissioner.
New regulation of internet search engines
It has only recently become clearer to the public how extensive the powers of the eSafety Commissioner are, and how easily they can be expanded. Quite separate from the enactment of the under 16s social media stipulation already discussed was a measure that introduced a requirement for internet search engines to meet new age verification standards from 27 December 2025, again raising questions about whether this will require all Australian internet users to register for a digital ID. This was not the result of legislation laid before Parliament (as the social media minimum age requirement was), but a new code registered under the Online Safety Act,[32] which eSafety will be responsible for monitoring, and enforcing. Senator Alex Antic characterised this development as the “expanding surveillance State”.[33]
Google and Microsoft are still working on how they can implement it. One option noted by Senator Antic, to avoid the potential for systematic over-censorship, is to follow the model adopted by Texas and Louisiana. Instead of placing checkpoints across the entire internet, those jurisdictions were able to focus on particular websites which specialise in relevant harmful material. The privacy of Australians would be impacted minimally by that approach.
This raises classic rule of law questions of proportionality propounded by the ancients (Cicero, Justinian, Augustine, and Grotius). Actions should be commensurate with the goals they aim to achieve.[34] Fundamental human rights principles similarly demand that the least intrusive means of achieving a legitimate restriction must be chosen. In the context of this particular measure it means it should only be imposed where necessary and provided it will be effective. Claims that it would not be effective should therefore be seriously evaluated (for example, that it may not even do much to protect young people),[35] and more general observations should be heeded, like that from Lisa Given, Professor of Information Sciences at RMIT, that “I have not seen anything like this anywhere else in the world”.[36]
Hate speech, vilification and similar prohibitions
It is a basic rule of law requirement that the law be readily understood, clearly stated and predictable. It is at risk of being compromised by the gap between the harm-based standards set by the Online Safety Act 2021 and those applied by the eSafety Commissioner, as recent proceedings brought by Chris Elston (known as ‘Billboard Chris’) show. Billboard Chris challenged the eSafety Commissioner’s order to X Corp to ‘take down’ one of his posts that was critical of transgender activist Teddy Cook. Interpreting the relevant provisions of the Act, the Administrative Review Tribunal expressed the view that “Parliament was not seeking to control or regulate debate on controversial issues, nor to manage or set minimum standards of courtesy as to how such debates should be conducted” (notwithstanding the impression given by claims that content deserves to be removed).[37] Instead, these provisions “were designed to protect individuals from specific online attacks that can reasonably be understood as having the intention of causing serious harm to them in particular” (in the sense of serious distress, physical harm or harm to mental health).[38]
The eSafety Commissioner is aware that hate speech is popularly misunderstood as occurring at an inappropriately low threshold[39] and includes “anything negative that was directed at another person”, which is much lower even than what simply causes offence.[40] No official steps appear to have been taken to correct that misapprehension. It is important to do so given such misunderstandings are likely to produce user complaints at content that is not in any real sense harmful but flagged for removal, nonetheless. Also, if the public perception is that such a low threshold applies under hate speech and related legislation, it results in unjustified self-censorship on a considerable scale.
The eSafety Commissioner can take action against content that is unlawful for being discriminatory or vilifying. It must meet a specific legal threshold for adult cyber abuse to trigger powers under the Online Safety Act, including that an individual was targeted. At state and territory level, discrimination and vilification provisions are the subject of numerous reviews by Labor governments (e.g., recently in Queensland and Victoria, and currently in New South Wales) and remain responsive to pressure to expand the prohibitions, especially so they operate at a lower threshold. Existing vilification provisions in Tasmania and the Northern Territory are very readily triggered, by conduct likely to cause offence or insult (among other things), even though no harm need be proved. Since provisions like these can be used to shut down viewpoint difference on wide-ranging subjects that touch on any protected attribute, they can be very effective in producing self-censorship on a large scale, in themselves, exacerbated by the eSafety Commissioner’s powers.
In provisions that apply at such a low threshold, the rule of law requirement does not appear to be met, that laws must be clear, certain, and predictable.
The eSafety Commissioner has broad powers of censorship, exercised in an environment that is opaque, involving complex interactions with digital service providers, in which at least three differing standards for content regulation are meant to converge, but do not. These are set by:
- the Online Safety Act (the statutory threshold)
- the eSafety Commissioner (in published guides, and according to practice which has been shown to depart from the statutory standard),
- and in the user terms and conditions of the service providers.
Compounding the mysteries of the content curation process are the fact-checking and other procedures followed by service providers, and the algorithms applied. It remains to be seen whether the eSafety Commissioner’s powers will be expanded further in light of the current review of the Online Safety Act that was scheduled at the time the Act was passed.[41]
Privacy is given primacy
Whatever the Billboard Chris case may have done to recalibrate the eSafety Commissioner to the censorship standards of the Online Safety Act, some consider this may have been nullified by legislative changes in some Australian states as they relate to privacy.[42] Recent changes at federal level to privacy laws have a major impact.[43]
Privacy laws were certainly due for reform, since statutory privacy protection before then had been confined rather ineffectually to regulating the way in which personal information is managed by governmental agencies (and certain private sector entities).[44] It was necessary to address the escalation of fraud and doxxing. However, the new privacy legislation not only far exceeded the protection that was needed to protect privacy (under the ICCPR), it did so at avoidable and substantial cost to the freedom of expression.
It created new criminal offences, for disseminating personal data in a way that is menacing or harassing, and for targeting an individual or group because of a distinguishing attribute (such as race, religion, sex, sexual orientation, gender identity, intersex status, disability, nationality or national or ethnic origin). Among the concerns expressed are that these provisions do not correspond with what most people think is doxxing. The Free Speech Union pointed out that it includes misgendering, dead-naming, and even pointing out which societies or organisations (such as golf clubs) politicians frequent, and that much of the really harmful conduct is already criminalised (such as physical threats, stalking, and harassment).[45] The Law Council of Australia was especially concerned that the doxxing provisions prohibited legitimate public interest journalism exposing contradictory, unethical, or illegal behaviour by public officials or businesspeople.[46]
The new legislation also created a statutory tort of privacy. Such a tort is not recognised in most jurisdictions (only a few, such as New Zealand, the US and Canada). However, the new Australian one is very different in totality. It creates civil liability for “serious invasions of privacy”, in circumstances where there is a “reasonable expectation of privacy”. This is determined, unusually, by “a person in the position of the plaintiff”, i.e., the person standing to gain financially from making a claim. No proof is needed of damage (which may be psychological). We can expect the new privacy tort to strongly encourage litigation. This too might meet the description of being like nothing “anywhere else in the world”.
A real issue, as already noted, is that the recognition in law is lacking in Australia for freedom of expression in the ICCPR sense. The bill was criticised in consultation for giving primacy to the right to privacy, ignoring the right of freedom of expression that is also enshrined in the ICCPR.[47] In practice, the new privacy legislation is likely to operate principally by fear, producing self-censorship to avoid activist litigation. Instead of freedom of expression being preserved (which it should be in any privacy legislation purporting to implement treaty standards — by respecting the proper parameters of privacy and freedom of expression) the freedom of expression operates merely as one of many considerations in the mix when determining if a countervailing public interest should prevent a successful private action for damages. The legitimate exercise of freedom of expression provides no defence to either the civil or criminal provisions, which is a clear indication that the settings in the legislation are seriously wrong. The freedom is given no real place, even where it should have full force and operation in Australian law. This is contrary to the Australian Law Reform Commission’s 2014 conclusion on reforms to privacy law that “privacy should be balanced with other rights and interests, such as freedom of expression”.[48]
Hate crimes with no relief for political speech
It is understandable in the aftermath of 7 October 2023, and with the religious tension that followed in Australia, that the government would want to step up its response to religiously motivated hate crimes. However, the amendments to Commonwealth criminal legislation that were passed in late February 2025 were very unusual.[49] While it is quite normal for hate speech legislation to prohibit seriously harmful speech directed at those with particular protected attributes (in this case corresponding with attributes protected by discrimination),[50] it is novel in Commonwealth criminal legislation to apply liability at the threshold of whether a “reasonable member of the targeted group would fear that the threat” of force or violence would be carried out. It lowers the threshold for criminal liability considerably.
It also provokes an important rule of law issue, that the law must be foreseeable in its consequences. Rather than adopt a more objective test (e.g., involving a hypothetical reasonable person in the community at large, rather than possibly a small minority group), the new law determines criminality on the basis of a reaction of ‘fear’ that results from what was said or done, within someone whose own life experiences may not be familiar to many. This is likely to produce excessive self-censorship, especially in light of the maximum prison sentences which the offences attract (five years and seven years, depending on the offence).
Among those raising objections were the Parliamentary Joint Committee on Human Rights (PJCHR) with oversight of human rights issues, and the Law Council of Australia, criticising the serious adverse impact on freedom of expression as a result of the combination of several factors. Many new provisions were the result of late-stage amendments without full opportunity for reflection. The text of the bill was voluminous and the implications not obvious on the face of the text.
One important change occurred as a result of text which simply read: “Subsection 80.3(1) of the Criminal Code: After ‘and C’, insert ‘(other than sections 80.2A, 80.2B, 80.2BA and 16 80.2BB)’.” Only those willing to invest time in researching what this meant would appreciate its profundity. The effect was to remove an age-old defence to protect a person who in good faith was caught by the Criminal Code provisions while engaging in the very things that should be welcome and supported in a democratic society, and are envisaged by many rule of law principles.
These include (according to the defence that has now been removed) where a person: “points out in good faith errors or defects in [government, legislation or the administration of justice], with a view to reforming those errors or defects”; “urges in good faith another person to attempt to lawfully procure a change to any matter established by law, policy or practice in the Commonwealth, a State, a Territory or another country”; “points out in good faith any matters that are producing, or have a tendency to produce, feelings of ill‑will or hostility between different groups, in order to bring about the removal of those matters”; or “publishes in good faith a report or commentary about a matter of public interest”.[51]
The government’s reason for removing the defence was that violence on the basis of a person’s protected attributes can never be done in ‘good faith’. It sounds compelling but is not an adequate answer to the remarkable fact that such ordinarily legitimate acts of free speech and democratic engagement as were covered by the defence could now make criminals of Australians. In light of that, the government’s rationale for removing the defence sounds more canted than convincing.
By various means the legislation considerably lowered the threshold at which a person becomes criminally liable, applying concepts that are not as clear as one would wish. If the terms ‘threatening’ and ‘force or violence’ sound clear enough, consider the eSafety Commissioner’s official understanding of ‘violence’. It includes non-physical violence or abuse because of biased or harmful beliefs.[52]
The Parliamentary Joint Committee on Human Rights in its scrutiny function pointed out that there was “a risk that, in practice, the offences could capture a greater range of conduct that may be offensive and insulting but the prohibition of which may constitute an impermissible limit on the rights to freedom of expression and religion. In relation to the offences related to urging violence, this risk appears to be particularly pronounced given the proposed removal of the defence of acting in good faith”.[53] The Law Council of Australia recommended that “the entire division [of the legislation] be reviewed afresh by the Australian Law Reform Commission in light of recent developments and the important fundamental freedoms and human rights at stake”.[54]
The effect of legislation that curtails speech at too low a threshold can be (if taken too far) to protect ideologies and prevailing narratives, which in principle should be open to legitimate debate.
The proper functions of Parliament compromised
The lawmaking function of Parliament is a key aspect of the rule of law, both in the processes by which legislation is created (publicly promulgated, with due consideration by Parliament, with legislative scrutiny, and democratic accountability) and in the properties of the resulting law (including, as already noted, that it should be certain and predictable and comport with international human rights standards). It appeared, particularly towards the end of 2024, that some aspects of the ordinary parliamentary process were unusually compromised.
The politicisation of the privacy bill is telling in the comments of the Deputy-Chair of the Legal and Constitutional Affairs Legislation Committee: “the bill deals with a number of discrete policy areas where: (a) it could be reasonably anticipated that there would be divergent views in the Parliament; and (b) different scrutiny concerns are raised … By including three schedules dealing with disparate matters in a single bill, senators will be forced to decide whether or not to support the bill in circumstances where they may agree to one or two schedules, but not all schedules. This is not best practice (if the goal is not to ‘wedge’ political opponents)”.[55]
The scrutiny of bills was also severely truncated. The same Deputy-Chair also objected in relation to the privacy bill “that inappropriately abbreviated timelines place a great workload upon the secretariat; especially when multiple bills dealing with important matters are required to be dealt with contemporaneously”.[56] The same point was echoed by the Law Council, observing that on five previous occasions complex legislation was being presented with inadequate time even for legal experts to get their heads around.[57]
The parliamentary inquiry report for the social media minimum age bill was tabled on 18 November 2024 and later the same week (on 21 November), the bill giving effect to it was introduced to Parliament. The closing date for public consultation was 22 November! The bill was referred to the relevant committee for report by 26 November.
In one of the most chaotic sitting days imaginable, 32 bills passed through the Senate on 27 November 2024, including both the privacy and social media minimum age bills. A guillotine motion radically curtailing debate initially failed but was eventually passed, greatly aided by Greens support secured by Labor promising $500 million for social home improvements and the cancellation of support for coal, oil and gas investments and programs. In the course of the limited debate then allowed, Liberal Senator Richard Colbeck commented: “This government shows its disrespect for the parliament by trying to shove [the social media minimum age bill] through in the way it has”.[58] Senator Alex Antic objected to the truncated debate, that this is “not the way to run a democracy”.[59] The Greens tried to persuade the government to allow for proper consultation and scrutiny of that bill. Greens Senator Sarah Hanson-Young complained the bill was rushed and flawed, and does not do what it claims to do, in making social media safer for young people.[60] The misinformation bill was abandoned only a few days before 27 November, in all likelihood to give safe passage to these other 30 or more bills, including seriously controversial ones.
The conduct of parliamentary inquiries
Parliamentary inquiries form an important scrutiny and accountability role in support of the rule of law. The terms of reference of December 2023 for the Inquiry into right wing extremist movements in Australia are unusual for a number of reasons. First, by exclusively targeting problematic sources of extremism only on the right (i.e., “persons holding extremist right-wing views”, “right-wing extremist movements”, and “right-wing extremism”), the inquiry would avoid criticism of those on the political left (an asymmetry that Australian intelligence and law enforcement agencies did not find helpful). Such imbalance itself has potential to promote a narrative that somehow right-wing extremism is inherently more dangerous than left-wing extremism. It would also tend to invite outcomes in law which do not apply equally to everyone, among other rule of law implications. Secondly, although the terms of reference made a solitary reference to violent extremism the inquiry is far broader. This matters, because there is a propensity, recognised within the UN, for the official condemnation of non-violent activities said to relate to ‘extremism’ to illegitimately quell free speech.[61] It is also notable that the terms of reference cover activities attributed with mere ‘capacity’ for violence or which may be said to form part of an indirect progression towards violence, with origins as remote as mere ‘views’.
The inquiry appears to have undergone something of a positive course correction between the terms of reference and the report, in conclusions and recommendations not confined only to right-wing extremism. However, the potential for the recommendations to give rise to outcomes which unjustifiably restrict free speech is still real. Free speech concerns are triggered, not at that end of the spectrum where violence and other concrete harms are most obvious (such as when individuals or groups use fear, terror, or violence to further or achieve ideological aims), but where there is no real risk of harm at all. That is the domain where democratic freedoms face greatest risk, such as selective enforcement, even by measures that on their face apply generally.
The inquiry report mentioned Australian ‘values’ quite extensively, and acknowledged that “Australia is a healthy and vibrant democracy” in which “[f]/reedom of speech is fundamental to Australia’s values”[62] — although it did not go into the reality that the free speech values are not well reflected in substantive law. In its conclusions and recommendations the report at times maintained the critical connection with violence (such as when commenting that “[t]he threat or use of violence against specific groups of people is an attack against our shared values”);[63] or that “[p]/olitical views become unacceptable when individuals or groups use fear, terror, or violence to further or achieve ideological aims”. However, that vital connection, and indeed any connection with conventional criminal acts, is absent in important places, such as when making the vague assertion that “extremist movements pose a threat to Australian society and Australian values”. ‘Values’ language without proper constraint could lead to elastic interpretation, capable of rendering ordinary attitudes extreme, including those based on traditional or religious beliefs without even the remotest connection with physical violence or crime. Parallels with the UK are not encouraging. The UK Christian Institute warns against values-based stipulations, apparently sold to the public as a solution to extremism, which it described as “a tool for promoting political correctness”,[64] and “a woke crusade”.[65]
In its conclusions and recommendations, the inquiry report covered situations unconnected with violence, including that “Australians who belong to minority groups reported the feeling of alienation and exclusion associated with being targeted by extremists. Their evidence demonstrated that even non-violent actions can have severe consequences for their sense of belonging and participation in society”.[66] At this point the report seems to appeal to a solution in criminal hate speech legislation of exactly the type accomplished by the government in February 2025 (see ‘hate crimes’ above), which was enacted after the inquiry report was published. It would also be of concern if the recommendations resulted in yet more hate speech legislation or expanded powers of the eSafety Commissioner, given the liberality with which online content has been removed under the eSafety Commissioner’s existing powers.
Conclusion
Do rule of law principles need to be reclaimed, and if so how?
A certain commonality emerges across the issues traversed in this article, concerning the posture of government in a strongly protective role, against the harms posed by misinformation, online content, invasions of privacy, hate speech, and extremism. There are genuine harms that need to be addressed but legislative and other responses by any government are susceptible to being misplaced or disproportionate. At the heart of recent governmental action appears to be a struggle to control public discourse, by legislative and other mechanisms, in an age when knowledge is proliferating like never before and access to wide-ranging perspectives is met with official alarm. Concepts like free speech, perceptions of truth and viewpoint diversity are undervalued in such contention.
Measures authorising shutdown and censorship seem to be increasingly the go-to response to protect against a greater range of harm than ever before. Some of the examples discussed here in certain respects suffer from the defect that their ostensible aim is not even fulfilled (e.g., social media minimum age legislation), or they are disproportionate to the point that they prompt speculation that the government is willing to curtail political expression, other harmless forms of expression, or accountability (misinformation, hate crimes and privacy legislation). Some measures rely on an expansion of ‘harms’ without necessarily a compelling rationale or evidence base sufficient to assure us that those harms are real, not merely hypothetical (extension of age-related regulation to YouTube). It may legitimately be asked what then is the intended purpose of such excessive restriction?
A companion article to follow this paper, on the concept of ‘legitimacy’, explores much further the possessive nature of some governments in their protective role. It seeks to show that the modelling of the role of the State in France, from the philosophical underpinnings of the revolution through to the present day, strongly favours central power, and strongly protective government in the service of collective interests, in which the freedoms of the individual are subordinated. This contrasts with the position in America, where the colonies won their independence by resistance against unwanted State intrusion, and where a high value is placed on individual freedoms, especially free speech, to this day. Such a comparison can be useful in understanding how a government, even in Australia, perceives its role.
The rule of law is protective of the proper functioning of a democratic society. Yet the examples discussed in this article seem to demonstrate the potential for rule of law principles to be sacrificed by an over-protective function assumed by government.
The issues canvassed here were selected to illustrate how the rule of law is intended to infuse the conduct and output of government, and promote the fundamental rights of all Australians in the process. The rule of law is meant to be more than just theoretical. It points to the need for tools of accountability and transparency so that power is exercised fairly and responsibly. It shows how and why laws should be the proper product of due parliamentary consideration, be clear, predictable and uniform in their application, and give full support for all fundamental human rights for everyone, without favouring some over others. It indicates limits to how government may pursue its policies and mandate.
However, to be effective, rule of law principles need to be embedded within the culture and operations of all branches of government. They cannot be assumed to be fulfilled in Australia simply because of the existence of certain constitutional checks and balances, parliamentary processes and conventions, and statutory and common law sources that support human rights and due process. These are incomplete and in practice inadequate to the task.
If there is material slippage in the practical operation of the rule of law it is vital to avoid any habituation of that state, or further decline. There are no systems in place to re-establish what is lost by ignoring rule of law principles. As a last resort there is the innate recognition, or reflex, that warns that things are not what they should be, triggered like a phrenic nerve that tells us when we are not breathing. It is too pessimistic to think we have reached that point. Rule of law principles point specifically to what needs to be recovered, and we have democratic and other freedoms sufficient, though imperfect, to engage to that end.
Dr Paul M Taylor is an Honorary Senior Lecturer at the TC Beirne School of Law, University of Queensland, and Fellow of the Centre for Public, International and Comparative Law; an Adjunct Professor at the School of Law, University of Notre Dame, Sydney, Australia; and Fellow of the Higher Education Academy. However, the views expressed in this article are not attributed to any institution with which he is affiliated.
Endnotes
[1] Sir David Williams lecture at Cambridge University Law Faculty, 2006.
[2] Bingham, T. (2010). The Rule of Law, Penguin Books. Others might express these principles in terms of equality before the law, supremacy of law, fairness and predictability, an independent judiciary, separation of powers, transparency and accountability of government, and access to justice.
[3] Bingham, T., ‘The Rule of Law in the International Legal Order’, in McCorquodale, R. (ed.) (2010), The rule of law in international and comparative context, p3.
[4] Ius is a legal constitutionalism concept expressed in different ways e.g. by Sandro, P. (2022) The Making of Constitutional Democracy. Hart Publishing; Kotowski, A. (2023) Ius or Lex: The evolution of the concept of law in Continental Europe.
[5] Waldron, J. (2016) ‘The Rule of Law’ in Stanford Encyclopedia of Philosophy (22 June 2016). Stanford University.
[6] Security Council, The rule of law and transitional justice in conflict and post-conflict societies, S/2004/616, para 6.
[7], UN Secretary-General, A New Vision for the Rule of Law, 2023 (follow up to ‘Our Common Agenda’ (A/75/982) 2021, p.4.
[8] Law Council of Australia, Policy Statement, Rule of Law Principles, March 2011.
[9] Aristotle, ‘Politics’, 1252b27-30.
[10] Bielefeldt, H. (2020) ’Limiting Permissible Limitations: How to Preserve the Substance of Religious Freedom, Religion and Human Rights’, Religion and Human Rights 15, 3-19), p.13.
[11] ICCPR article 2.
[12] The implied freedom allows Australians “to exercise a free and informed choice as electors” (Lange v Australian Broadcasting Corporation (1997) 189 CLR 520, 570).
[13] See Ravbar v Commonwealth of Australia [2025] HCA 25 (18 June 2025), per Steward J. at [270 – 296] challenging the existence of the implied freedom.
[14] Lange v Australian Broadcasting Corporation, 564.
[15] Communications Legislation Amendment (Combatting Misinformation and Disinformation) Bills, 2023 and 2024.
[16] Sen. Alex Antic, Home Affairs Freedom of Information Request, 24 May 2023 (https://www.alexantic.com.au/home_affairs_freedom_of_information_request).
[17] Sky News, ‘Twitter files reveal extent of Australian government’s online COVID censorship’, 25 May 2023 (https://www.skynews.com.au/opinion/chris-kenny/twitter-files-reveal-extent-of-australian-governments-online-covid-censorship/video/d4f88176f380145ce80cf9315bd00140).
[18] The Online Safety (Basic Online Safety Expectations) Amendment Determination 2024 amended the Online Safety (Basic Online Safety Expectations) Determination 2022 made under Part 4 of the Online Safety Act 2021. It expanded the basic online safety expectations of digital service providers.
[19] Explanatory Memorandum, p.39.
[20] Lange v Australian Broadcasting Corporation, 570. Although it can cover a potentially wide range of different forms of expression, it has limited application as a freedom. It only operates as a constitutional limit on legislation that unduly restricts the free flow of information and discussion about government and political matters, which is considered essential for a functioning representative democracy.
[21] Explanatory Memorandum, p.1
[22].Prof. Anne Twomey, ‘Is the “ban” on under-16s from YouTube constitutionally valid?’, https://www.youtube.com/watch?v=7aC_Ka8_7u4.
[23] Explanatory Memorandum, p.1
[24] Enhancing Online Safety for Children Act 2015.
[25] Klein, J. ‘World Economic Forum: Australia’s eSafety Commissioner Calls for ‘Recalibration’ of Human Rights, Free Speech,’ 24 May 2022.
[26] e-Safety Commissioner v X Corp, [2024] FCA 499; 303 FCR 354.
[27] Baumgarten v eSafety Commissioner, [2025] ARTA 59.
[28] ‘eSafety Commissioner news – When is a non-notice a notice?’, Australian Law in Practice 2025, 13 February 2025 (https://adminlaw.com.au/2025/02/13/when-is-a-non-notice-a-formal-notice-under-an-act-and-the-dangers-of-avoiding-statutory-steps/).
[29] Jesmine Cheong & Veronica Lenard, SBS News, 24 June 2025. (https://www.sbs.com.au/news/article/digital-swimming-lessons-five-takeaways-from-the-esafety-commissioners-press-club-address/lsonrtmy0).
[30] Darren Davidson, ‘ “Like banning a library”: eSafety Commissioner “misled” Australians on push to ban YouTube’, Sky News Australia, 11 July 2025.
[31] Sky News, ibid.
[32] Schedule 3 – Internet Search Engine Services Online Safety Code (Class 1C and Class 2 Material) (Code) registered under the Online Safety Act 2021 (Cth) on 27 June 2025, with an effective date of 27 December 2025.
[33] Alex Antic, ‘The Expanding Surveillance State’, YouTube, 17 July 2025 (https://www.youtube.com/watch?v=zxHtX1EqTzA).
[34] Engle, E. (2009)’The history of the general principle of proportionality: an overview’, Dartmouth Law Journal, Vol. X.:1, 2009, pp. 2-4.
[35] Ange Lavoipierre, ‘Australia is quietly introducing ‘unprecedented’ age checks for search engines like Google’, ABC News, 11 July 2025 (https://www.abc.net.au/news/2025-07-11/age-verification-search-engines/105516256).
[36] Ibid.
[37] Christopher Elston v eSafety Commissioner, 2024/2583, ART Decision, 1 July 2025, para. 192.
[38] Ibid.
[39] Report, Online Hate Speech: Findings from Australia, New Zealand and Europe (https://www.esafety.gov.au/research/online-hate-speech).
[40] Ibid, p.7.
[41] Statutory Review of the Online Safety Act 2021 (https://www.infrastructure.gov.au/have-your-say/statutory-review-online-safety-act-2021).
[42] Deves, K. ‘Legislating identity: the erosion of sex-based rights and a free-speech victory that fought back’, The Spectator Australia, July 2025, p.3.
[43] The Privacy and Other Legislation Amendment Bill 2024 was passed on 29 November 2024.
[44] The Privacy Act 1988 (Cth).
[45] Free Speech Union of Australia, Submission 4, pp. 1–2.
[46] Law Council of Australia, Submission 67, p.38.
[47] Legal and Constitutional Affairs Legislation Committee Report, November 2024, p.55.
[48] ALRC Report 123, ‘Serious Invasions of Privacy in the Digital Era’, 2014, p. 18.
[49] The Criminal Code Amendment (Hate Crimes) Act 2024.
[50] Race, religion, sex, 1 sexual orientation, gender identity, intersex status, disability, 2 nationality, national or ethnic origin or political opinion.
[51] Former s.80.3(1) of the Criminal Code.
[52] https://www.esafety.gov.au/key-topics/gendered-violence.
[53] Parliamentary Joint Committee on Human Rights, Report 9 of 2024, [1.215].
[54] Law Council of Australia, submission, [10].
[55] Report of the Legal and Constitutional Affairs Legislation Committee, Privacy and Other Legislation Amendment Bill 2024 [Provisions], p.104.
[56] Legal and Constitutional Affairs Legislation Committee, Privacy and Other Legislation Amendment Bill 2024 [Provisions], Report, November 2024, p 103.
[57] Law Council of Australia (Law Council), Submission 67; Committee report p. 105.
[58] Courtney Gould, Tessa Flemming and Tom Crowley, ‘Senate passes social media ban for under 16s among 32 bills — as it happened’, ABC News, 28 November 2024.
(https://www.abc.net.au/news/2024-11-28/federal-politics-live-november-28/104654902).
[59] Ibid.
[60] Ibid.
[61] Report, Special Rapporteur on the promotion and protection of human rights and fundamental freedoms, A/HRC/31/65 (2016), paras. 27 & 35.
[62] Report, Committee views and recommendations, para7.1,
[63] Ibid, para 7.1.
[64] ‘Ofsted and “British values” ‘, The Christian Institute, June 2017.
[65] ‘New extremism definition could pose threat to religious freedom’, The Christian Institute, 3 November 2023.
[66] Ibid, para 7.24.
