By orders made in July, three cases have been referred for the opinion of the Full Court of the High Court of Australia. Each of these cases concern the so-called implied freedom of political communication guaranteed by the Constitution, as authoritatively stated in Lange v Australian Broadcasting Corporation (1997) 189 CLR 520. In Lange, and each case that has followed it, the Court has said it is necessary to imply this freedom, not explicitly recognised in the terms of the Constitution themselves, to enable the system of responsible government established by the Constitution to function. The freedom operates to invalidate a law, including State law, that burdens it in a way that is not reasonably and appropriated adapted to serve a legitimate end the fulfilment of which is compatible with the maintenance of representative and responsible government. On account of these two features (that the right arises from ‘necessary’ implication, and it cuts down laws that a Court judges to be disproportionate), its application have proved difficult and inconsistent. These are cases to watch for anyone who has been following its jurisprudence.
The three cases are:
The first two cases relate to Pt 4A of the Online Safety Act 2021 (Cth). Section 63A of that Act summarises the operation of Part 4A as follows. Providers of certain kinds of social media platforms must take reasonable steps to prevent children who have not reached a minimum age from having accounts. Section 5 of the Act defines age-restricted user to mean an Australian child who has not reached 16 years. Reddit, Inc is a social media provider subject to the requirement to take reasonable steps to prevent age-restricted users from having accounts; The Digital Freedom Project Incorporated is an incorporated association whose objects include advocating for the rights of youth, who have not yet attained adulthood, to use and communicate through social media technologies and defending freedom of expression in digital contexts.
The third case relates to Part 5.3B of the Criminal Code (Cth). This part was inserted by the Combating Antisemitism Hate and Extremism (Criminal and Migration Laws) Act 2026 (Cth), enacted after the terrorist attack at Bondi Beach on 14 December 2025. Division 114B of Part 5.3B makes it an offence for a person to be a member of a prohibited hate group, and creates several other offences for a persons involved in the activities of such a group—this includes activities to support the continued existence of the organisation. A prohibited hate group is an organisation that has been specified in the regulations. The Governor-General may specify an organisation as a prohibited hate group if the AFP Minister is satisfied that
the organisation has engaged in, or has planned to engage in, conduct constituting a hate crime; or has advocated engaging in conduct constituting a hate crime; and
specifying the organisation is reasonably necessary to protect the Australian community or part of the Australian community against one or more certain types of harm (such as social, economic, psychological or physical harm).
The Minister is expressly not required to observe any requirements of procedural fairness in this decision. But he or she must seek the agreement from the Attorney-General, and arrange for the opposition leader to be briefed.
A hate crime is conduct that would amount to an offence related to inciting violence against a person or group of persons because of the race, national or ethnic origin of that person under Commonwealth, State or Territory law. It is also conduct, or threat of conduct, that involves, or would involve, any one of causing serious harm to a person or damage to property, causing a person’s death, endangering a person’s life, or creating a serious risk to the health or safety of a section of the public; and the conduct was engaged, or threatened, because the perpetrator believed the person is distinguished by race or national or ethnic origin, or property is associated with a person so distinguished. A hate crime includes conduct engaged before or after the section commences, and a person need not have been convicted a hate crime for an organisation to be declared a prohibited hate group.
According to the plaintiffs' submissions, the principal plaintiff is a political party and a voluntary association with assets registered under the Associations Incorporation Reform Act 2012 (Vic); it has around 1700 members. The association’s objects are to preserve and advance the cultural, historical and ethnic interests of Australians of European heritage; and to advocate certain immigration and population policies. In these submissions, the plaintiffs recognise the connection of the name of the association with the policy of the Immigration Restriction Act 1901 (Cth), colloquially referred to as the “White Australia policy”. On 16 May 2026, this plaintiff was specified as a prohibited hate group by the Criminal Code (Prohibited Hate Group—White Australia) Regulations 2026 (Cth).
The impugned laws in this case are not only challenged on the grounds that they impermissibly burden the implied freedom of political communication; they are also challenged on the basis that they are not supported by any legislative power of the Commonwealth, and they infringe Ch III of the Constitution (that is, the laws are invalid because they confer on the Commonwealth Executive the power to adjudicate and inflict punishment; the Constitution reposes this power in Chapter III Courts alone). Accordingly, the question as to whether the laws are invalid due to the implied freedom of political communication might not necessarily be answered.
There are three things in particular to watch in these cases. The first, in connection with the Reddit, Inc and The Digital Freedom Project Incorporated cases, is whether any real attention is paid to the first limb in Lange; namely, whether the law in operation effectively burden freedom of communication about government or political matters. This limb, more often than not, receives little attention in the reports; seemingly, there is very little that does not effectively burden the freedom of political communication. In these two cases, however, the Court has an opportunity to meaningfully engage with this limb given the entirety of the demographic affected by the changes to the Online Safety Act 2021 (Cth) do not directly participate in the system of representative and responsible government; in fact, there are very few countries where the political franchise is extended to persons as young as 16; there are none where it is extended any younger. And if we are to understand the system of representative government in light of the history and institutions of constitutional government in Great Britain before federation, can it be said that children have even a vicarious role in this system? The question may be put another way: does it effectively burden the freedom of political communication guaranteed by the Constitution if the most prominent social media websites are not legally accessible to children, who cannot participate directly in that system, because they may or may not use those websites to discuss political matters with others who may or may not participate directly in that system of government, or may no longer have access to information contained on those websites that may pertain to Australian politics? Recalling that to meet this test it is not simply does the law impose a burden, but does the law impose an effective burden, I would submit that the answer is not so simple as to be assumed.
The second, in connection with all three cases, is whether there develops any greater consistency in approaching the second limb of Lange—the question whether the law is reasonably appropriate and adapted to serve a legitimate end the fulfilment of which is compatible with the maintenance of the constitutionally prescribed system of representative and responsible government. This lack of consistency is said to emerge from the observations made in the first point; that is, the first limb is seemingly a nominal gateway. As a result, it to is the second limb that the work must be done to ensure the restriction on legislative power does not go too far so as to be unworkable. Many cases have tested the Court’s resolve in constraining the freedom’s impact; and in McCloy v New South Wales (2015) 257 CLR 178, the majority imported the test of ‘structured proportionality’ from the human rights jurisprudence of Europe and Canada. This approach asks three definite questions to determine whether a law is reasonably and appropriately adapted:
is the law suitable as having a rational connection to the purpose of the provision?
is the law necessary in the sense that there is no obvious and compelling alternative, reasonably practicable means of achieving the same purpose which has a less-restrictive effect on the freedom?
is the law adequate in its balance—a criterion requiring a value judgment consistently within the limits of the judicial function, describing the balance between the importance of the purposes served by the restrictive measure and the extent of the restriction it imposes of the freedom?
If the answer to any of these questions is ‘No’, then the law fails for lack of power.
The general consensus is that the principal advantage of this test is that it makes the reasoning for a decision transparent; however, it has been controversial since McCloy; indeed, in McCloy, Gageler J (now Chief Justice) and Gordon J, both members of the current High Court bench, criticised and declined to use it. In three 2025 cases, the High Court had further discussed the application of the structured proportionality test—these cases being Babet v Commonwealth; Palmer v Commonwealth [2025] HCA 21, Ravbar v Commonwealth [2025] HCA 25, and Farmer v Minister for Home Affairs [2025] HCA 38. In these cases, the controversy continued—sharply; but a majority of the bench found the test to be merely a ‘tool’ which may or may not be adopted among other tools. On the other hand, Edelman J and Steward J considered the structured proportionality test to be binding, as a result of the proceeding decisions; and that it would undermine the principal of stare decisis to depart from those cases without submissions from any party. As Dr Jemimah Roberts shrewdly recognises, two of the four of the aforementioned majority are the most recent appointees; this demonstrates the vulnerability of constitutional doctrine to changes to the composition of the bench. And as Edelman J noted in Ravbar, ‘there is presently no majority view as to when structured should be applied and, if it is not applied, which alternative should be deployed in those cases where a judge chooses not to apply structured proportionality’ and ‘the different formulations have “high costs to institutional integrity and legitimacy”.’
The third is the judgment of Steward J, should he deliver separate written reasons. This is all but certain. His Honour has consistently doubted the existence of the implication, most prominently in LibertyWorks Inc v The Commonwealtth (2021) 274 CLR 1, Babet, and Ravbar; and, in doing so, was rebuked in Ravbar, particularly by the Chief Justice (see [23]–[30]). His Honour’s reasoning is not simply a reiteration of points previously made by Dawson J, Callinan J and Heydon J; rather, it has had the benefit of the recent cases outlined above, which have emphasised the enduring issues present in the application of the freedom, and has introduced some new ones. The facts background to each of these cases, more than the other cases Steward J has sat on concerning the implied freedom, present irresistible opportunities to develop the attack further.
If it has not been clear from the tenor of this article, this author finds the existence of the implied freedom, as formulated and its current state, to be dubious, and the reasoning underlying its implication unconvincing. This is despite otherwise supporting the utility of freedom of speech on broader terms than those of the implied freedom. Worse still is its application—it simply operates too broadly, and the Court is forced into making value or policy judgements within an adversarial system of litigation which is simply unequipped facilitate such decision making—this made even worse by the import of the structured proportionality approach from Germany, whose Australian adaptation portrays a self-consciousness about these issues. The Court, naturally, jealously guards its functions reposed in it by Chapter III of the Constitution from the Executive and the Legislature, welding its ultimate power over these two branches of Government to fend of intrusions; but the Legislature, who, unlike United Kingdom’s Parliament, is subject to the law, does not have such power to defend incursions into its functions reposed by Chapter I. We have placed our trust in the Court itself to forebear from this. The weight of this trust cannot be overstated, for we have given this institution the ability to declare the operation and limits of legal power that cannot be undone with the legislative-making powers granted to our elected representatives. The Court, in its formulation of the implied freedom which has an inherent element of judging the value or appropriateness of a law made by a representative body, places itself in positions where its institutional credibility is at most risk. Of course, a Court must be willing and able to make deeply unpopular or controversial decisions where the law requires it. But it one thing to declare a law to be beyond power because, for instance, it has no real connection to any of a subject matter of power; or because, in the case of State legislation, it is invalid because it inconsistent with a federal law. It is another where the Court invalidates a law as being beyond power because its aims could achieved just as well in some alternative way, or because it is judged to have an imbalanced operation. It creates fertile ground for sowing a very American distrust in the motivations of the Justices of the Court, easily exploited by commentators read by laypersons. This distrust is misplaced, but it is a natural consequence of an interpretative approach to the Constitution, present to different degrees over the last 30 years, which is all too ready to find rights for the Court to vindicate, or restrictions on legislative power.