On 16 June 2021, the High Court delivered its judgment in LibertyWorks Inc v Commonwealth of Australia [2021] HCA 18 (LibertyWorks v Commonwealth). The case centred on the Foreign Influence Transparency Scheme Act 2018 (FITS Act), a legislative scheme introduced to expose foreign influence effected by foreign principals within Australia. In dispute were provisions which required certain foreign principals (in this case foreign political organisations), who communicate material to the Australian public, to register under the scheme. The primary question was whether the registration requirements impermissibly infringe the implied freedom of political communication (the implied freedom). The plaintiff, LibertyWorks, a free-speech and liberty advocating think tank, unsuccessfully argued that the impugned provisions infringed the implied freedom.

This case profiles, for the first time, the High Court’s approach to the FITS Act. In doing so, it provides an adequate, albeit not complete, opportunity to consider one of the core mechanisms through which foreign influence is regulated in Australia. Of course, this analysis is necessarily limited by the scope of the case, which focusses specifically on the application of the concept of communications activity (s 21 item 3), and the subsequent requirements that flow in light of the implied freedom. Broader questions, such as legislative and definitional breadth, as well as Secretarial discretion under the scheme remain, however.

This post accordingly anchors its analysis to LibertyWorks v Commonwealth. It provides brief background by way of the facts of that case and the legislative scheme. It then considers the judgment, discussing the majority and minority judgments in turn (including differing methods of analysing legislative validity), particularly focussing on the concern raised in the minority judgments regarding the creation of two repositories under the FITS Act. The post concludes by considering the remaining legislative breadth of the FITS Act and who registration obligations may apply to.


The case concerned a conference organised by the plaintiff in 2019 known as CPAC Australia. The plaintiff was assisted in organising CPAC Australia by the American Conservative Union (the ACU) an organisation not dissimilar to the plaintiff. The ACU organises the Conservative Political Action Conference (CPAC), which is held annually in the United States of America. In 2018, it was agreed that there would be collaboration between the plaintiff and the ACU in organising CPAC Australia. The ACU would provide the plaintiff with the names of potential speakers, including a host of American speakers, to ensure the event’s success.

In August 2019, the President of the plaintiff was contacted by the Commonwealth Attorney-General’s Department (the Department) regarding CPAC Australia. The plaintiff was alerted to the scheme of the FITS Act and advised that the ACU appeared to fall within the definition of a ‘foreign political organisation’ meaning the ACU could be considered a ‘foreign principal’. The plaintiff was asked to consider registering under the FITS Act, which was followed by the Department issuing the plaintiff with a formal notice. The plaintiff did not comply with the formal notice, contravening the offence provision of the FITS Act; the penalty for which is up to six months imprisonment.

On 7 February 2020, the plaintiff filed an action in the original jurisdiction of the High Court, arguing that the registration provisions of the FITS Act regarding communications activity by a person acting on behalf of a foreign principal burden the implied freedom, cannot be justified, and are therefore invalid.

The High Court, by a 5-2 majority, upheld the validity of the impugned provisions of the FITS Act. In particular, the majority held that s 21 item 3, with respect to the requirement to register ‘communications activities’ undertaken by a foreign principal in Australia ‘for the purpose of political or governmental influence’, was valid because the registration obligations do not infringe the implied freedom. The majority comprised Kiefel CJ, who wrote with Keane and Gleeson JJ (the joint judgment), and Edelman J and Steward J, who wrote separately. Justices Gageler and Gordon held that the impugned provisions of the FITS Act did infringe the implied freedom and were therefore invalid.

Legislative scheme

The landscape of growing concerns about foreign influence forms the backdrop for the specific tranche of Commonwealth legislation that includes the FITS Act. The FITS Act functions alongside the National Security Legislation Amendment (Espionage and Foreign Interference) Act 2018 (Cth) and the Electoral Legislation Amendment (Electoral Funding and Disclosure Reform) Act 2018 (Cth). Specifically, the object of the FITS Act is

to provide for a scheme for the registration of persons who undertake certain activities on behalf of foreign governments and other foreign principals, in order to improve the transparency of their activities on behalf of those foreign principals.

The suite of reforms was built upon four key pillars: sunlight, enforcement, deterrence and capability; with sunlight being ‘at the very centre’ (at 13148). The FITS Act, which was closely modelled on the United States Foreign Agents Registration Act (FARA), introduces a scheme ‘[t]o ensure activities are exposed to sunlight’ (at 13148, emphasis mine). In other words, the FITS Act is the legislative apparatus that aims to expose coercive, clandestine, or deceptive behaviour from foreign principals through a form of public disclosure. Such exposure is achieved, according to the FITS Act, by requiring certain individuals or entities who undertake certain activities on behalf of a foreign principal for the purpose of political or government influence to register on a public register.

Public registration results in relevant information being made available through a publicly accessible register. However, the public register does not necessarily contain all information provided to the Department. The Secretary is required to keep a separate register containing additional information, which is not publicly available (the difference between the two schemes is further discussed below in connection with the reasoning of the minority judges). At the time this post is published, there are 91 registrants, 205 foreign principals and 330 activities entered on the public register.

LibertyWorks v Commonwealth

In LibertyWorks v Commonwealth, registration obligations were contingent on whether the plaintiff was undertaking an activity on behalf of a foreign principal that is registrable (ss 20-23) under the FITS Act. The specific registrable activity it was said the plaintiff had undertaken was communications activity. A person undertakes communications activity if they communicate or distribute information to the public. Once a person registers under the FITS Act, they have certain responsibilities, which includes keeping records and providing disclosure of the foreign principal in question.

The plaintiff acknowledged in oral argument that the legislative definitions applied to its circumstances, accepting that CPAC Australia constituted a communications activity and acknowledging that the ACU is a foreign principal that exists primarily to pursue political objectives. The plaintiff therefore accepted that, subject to the question of legislative validity, it had registrable obligations under the FITS Act.

The primary question agreed to by the parties was whether the FITS Act is invalid to the extent that it imposes registration obligations with respect to communications activities on the ground that such obligations infringe the implied freedom. The plaintiff argued that the obligation of registration is likely to have ‘a chilling effect on people who just want to be part of the general discourse of matters of political interest and concern.’ Registration, they argued, therefore has a deterrent effect on political speech.

Majority judgment

The majority found that the impugned provisions regarding registration obligations were valid insofar as they did not infringe the implied freedom. In finding validity, the joint judgment undertook a structured proportionality assessment ([44]-[85]), with Edelman J (at [194]) supporting the joint judgment’s application of structured proportionality. Justice Steward (at [249]), while briefly suggesting that ‘it is arguable that the implied freedom does not exist’, nonetheless accepted that structured proportionality is an appropriate analytical tool to use in this case (at [247]).

The current formulation of structured proportionality analysis in Australia in relation to the implied freedom is predominantly derived from McCloy, with slight adaptations made in subsequent cases. In summary, the Court must consider three questions when determining whether a law impermissibly burdens the implied freedom. Firstly, does the law burden the freedom? If yes, is the purpose legitimate? If yes, is the law reasonably appropriate and adapted to advance the legitimate objective? In order to answer the third question, the Court considers three elements: suitability, necessity, and adequacy in balance (for an expanded discussion of the test, see most recently Murray Wesson (particularly pp. 5-8) on structured proportionality).

The majority Judges in LibertyWorks v Commonwealth all found that the impugned provisions burdened the implied freedom (joint judgment at [54]; Edelman J at [195]; and Steward J at [289]), with the joint judgment finding the burden of registration ‘modest’ (at [85]). Justice Edelman suggested that while the burden is ‘not excessive’ it did remain ‘substantial’ (at [194]). He concluded, however, that while the ‘burden imposed here might deter some… [it] does not prohibit…political communication’ (at [244]). The joint judgment highlighted that, had the plaintiff not entered an arrangement with the ACU, ‘it could have conducted the CPAC event without incurring an obligation to register’ (at [64]).

In terms of legislative purpose, the joint judgment (at [58]), with which Edelman J (at [194]) and Steward J (at [261]) agreed, held that the purpose of the FITS Act sought to

achieve transparency, in the sense of the exposure of foreign influence, as a means of preventing or minimising the risk that foreign governments or other foreign principals will exert influence on the integrity of Australia’s political or electoral processes, as has occurred elsewhere.

There was additional emphasis in the joint judgment that not only is that purpose legitimate as ‘consistent with the constitutionally prescribed system of representative government, it serves to protect it’ (at [61], emphasis mine). Such a purpose, they found, ‘may be a very important factor in the justification of a law’ (at [61]).

The joint judgment found that there is ‘a rational connection between the statutory purpose and the requirement of registration’ therefore finding it suitable (at [77]), with Edelman J (at [238]) and Steward J (at [291]) in agreement. The joint judgement also held that ‘[b]oth disclosure and registration are necessary for the achievement of the FITS Act’s purposes’ (at [84]), with Steward J (at [291]) in agreement. Justice Edelman, while arriving at the same conclusion, recognised that ‘[t]he test of reasonable necessity remains capable of further development and refinement’ (at [202]).

An interesting inclusion under the necessity limb of the analysis was discussion by the joint judgment suggesting that the registration requirement in the FITS Act allows for

members of the commentariat, such as journalists… [t]o ensure effective disclosure of the nature and extent of foreign interests at play in the affairs of this country that might otherwise remain undisclosed or dimly understood (at [83]).

The requirement of registration established by the FITS Act, they went on, ‘allows the commentariat to be alerted to the presence of foreign influencers in public affairs’ which ‘enables public debate to be informed in a way that would not be achieved by source disclosure to the recipients of a particular communication at the time of the communication’ (at [83]). This is an intriguing comment from the joint judgment in that it places importance on journalists providing transparency to the overall scheme. In doing so, it recognises—and arguably embeds—the role of the commentariat, such as journalists, in exposing foreign influence in the Australian democratic system.

Finally, in relation to adequacy in balance, the joint judgment held that the FITS Act has ‘a powerful public, protective purpose’, and ‘[t]hat important purpose cannot be said to be outweighed by a burden on the freedom which is modest’ (at [85]). Justice Edelman outlined his reluctance to find a law invalid at this stage. He suggested that finding a provision inadequate in balance ‘will often mean that Parliament is entirely precluded from achieving its legitimate policy objective’, which (in the context of a law burdening the implied freedom) would be a ‘remarkable outcome’ for ‘an implication founded upon representative democracy’ (at [201]).

Justice Steward was also cautious about the third limb in structured proportionality, relying on Edelman J’s judgment in Clubb v Edwards (at [493]) which suggested that in some other jurisdictions this limb has ‘been effectively abandoned’. Justice Steward suggested therefore that this limb ‘should… be approached with very considerable trepidation’ (at [293]).

Minority judgment

Justices Gageler and Gordon dissented in outcome, employing very similar analytical methodology to reach their respective conclusions. Both Gageler J (at [100]) and Gordon J (at [126]) found that the provisions burdened the implied freedom. Justice Gordon (at [184]) held, in relation to the second question (legitimate purpose), that the ‘impugned provisions go well beyond the legitimate purpose of the FITS Act’, and Gageler J (at [118]) found that the compulsion to be registered to engage in political communication ‘is not compatible with the constitutional freedom of political communication.’

In determining validity in relation to the third stage, Gageler J employed the ‘precedent-mandated Lange-Coleman-McCloy-Brown analysis’, followed by consideration in the third limb of whether the impugned law is

reasonably appropriate and adapted to advance that purpose in a manner compatible with the maintenance of the constitutionally prescribed system of government (at [93]).

Justice Gordon’s query at the third stage only slightly differs from Gageler J’s, whereby Gordon J (at [134]) asks (italicised difference from Gageler J) if the impugned provisions are

reasonably appropriate and adapted to advance that legitimate purpose in a manner consistent with the maintenance of the constitutionally prescribed system of government?

It does not appear that the minor difference in phrase presents any major differences in analysis or outcome.

Further, it is worth emphasising the firm defence of the structured proportionality approach that was emphasised in the joint judgment and in the judgment of Edelman J. For example, the joint judgment (at [48], emphasis mine) noted that structured proportionality as an analytical tool ‘has consistently been maintained by a majority of this Court in each of the cases concerning the implied freedom since McCloy’. Justice Edelman, perhaps providing a firmer defence of the approach, stated that

In contrast with a vague, ad hoc application which purports directly to apply the triple uncertainty in the phrase “reasonably appropriate and adapted”, a structured proportionality analysis provides a transparent manner in which to determine whether a law which burdens political communication for some legitimate purpose has contravened the implied freedom of political communication (at [200]).

While Gageler J does not employ the structured proportionality approach, he stated that he was nonetheless ‘[d]oing [his] best to express that incompatibility in the language of structured proportionality’ (at [119]). He found the impugned provisions not suitable, not necessary, and ‘at the very least’, not adequate in balance (at [119]) because of the two repositories discussed below. Thus, Gageler J, despite not finding structured proportionality useful as a tool, clearly considered himself required by precedent to consider its application. This is a softening in his resistance against structured proportionality compared to, for example, his stronger refutation in Clubb v Edwards (at [158]-[160]).

The minority’s findings relied on the incongruence both Judges found between the object of improving transparency and the registration requirements which result in, as both Judges highlight, two separate repositories of information.

The two repositories under the FITS Act Scheme

Both Gageler and Gordon JJ were concerned that the FITS Act establishes two separate repositories of information. One is a ‘register’ kept by the Secretary which is not made public and the other a publicly accessible website. The explanation of concern is offered by Gordon J:

[w]hat is not made public cannot “improve the transparency [the purpose of the Act]” of activities on behalf of a foreign principal. The gap between the two repositories is not justified and cannot be bridged (at [129]).

Justice Gordon continued that ‘[i]t is evident from the text of the FITS Act and the FITS Rules that the “register” and “website” are separate and do not mirror each other’, with the public website containing a more limited class of information (at [160]-[161]). She concluded, reflecting on the ‘sunlight’ core of the FITS Act, that ‘a non-public register does nothing to minimise the risk of undisclosed influence’, rather ‘[i]t does the opposite.’ A non-public register ‘is in darkness, not sunlight’ (at [130]).

Similarly, Gageler J referred to the private register as a ‘secret register’, stating that

A scheme of registration narrowly tailored to improve transparency of political communication undertaken on behalf of foreign principals with the public or sections of the public in a manner that minimally impaired freedom of political communication would have no place for a secret register at all (at [117]).

A secret register, whereby information could be shared only with Commonwealth, territory, and state agencies

burdens political communication by a registrant with the public or a section of the public to a substantially greater extent than is necessary to achieve the sole identified legislative object of improving the transparency of that communication (at [118]).

Justice Gageler held, for the above reasons, that the ‘compulsion to be registered under the scheme in order to engage in political communication… is not compatible with the constitutional freedom of political communication’ (at [118]).

It should be noted that there was additional concern raised by both Judges (Gageler J at [94]-[101]; Gordon J at [179]) about the notion of ‘prior restraint’ in relation to the implied freedom, which was refuted to some extent by Edelman J (at [219]). This deserves further consideration but for sake of brevity is not considered in this post.

Breadth of registration obligations remains for civil society

In LibertyWorks v Commonwealth the High Court examined specific provisions of the FITS Act, but the potential breadth of non-impugned registration obligations remains to be considered. Concerns about such breadth have previously been raised by Anne Twomey, for example, in her submission to the Parliamentary Joint Committee on Intelligence and Security when it considered the FITS Bill.

Some of the Judges in LibertyWorks v Commonwealth commented on the ways in which broad terminology might capture wide parts of civil society. Indeed, as outlined by Gordon J (at [140]), ‘what is “political” is undefined’. This may capture, as Gordon J (at [140]) recognised, foreign organisations that exist ‘primarily to pursue political objectives’ by way of capturing them as foreign principals. Other examples of breadth include the definitions of ‘person’ and ‘information or material’, with Gordon J noting (at [139]) that ‘person’ ‘would include “think-tanks” (like LibertyWorks), public interest groups, academic institutions, media organisations and individual citizens alike’. Justice Gordon further noted that ‘information or material’ (the communication or distribution of which constitutes ‘communications activity’) would include ‘academic work’.

Therein lies potential registration obligations for those in academia or in public interest civil society organisations. For example, the FITS Act could impose obligations on academics and researchers who liaise with international colleagues or bodies. On that basis, as Steward J (at [275]) suggested

an Australian academic who prepares a paper (that constitutes a communications activity for the purpose of political or governmental influence) under an arrangement or understanding (perhaps to deliver the paper at an international conference) with a foreign academic (who is a foreign principal) who proposes to prepare her or his own paper might be liable to be registered.

This was accepted by the Solicitor-General in oral argument and acknowledged by Steward J as a ‘necessary by-product of important legislation which is broad-based and neutral’ (at [275]). The University of Adelaide, for example, recognises potential obligations for academics under the FITS Act.

Justice Edelman recognised that ‘a conscious choice’ was made in not incorporating academics into the exemptions (ss 24-30). Justice Edelman (at [215]), therefore, also recognised the potential for the application of the FITS Act to the university and research sectors

The regulation of registrable communications activity might, therefore, extend to communications by academic researchers in Australia whose public research output is conducted with funding from any company in which more than 15% of the issued share capital is held by a foreign organisation that exists primarily to pursue political objectives. If the funding of those communications meant that they were undertaken ‘under an arrangement’ then they would be registrable communications activities if the academic had a substantial purpose to ‘affect in any way’ a section of the public, such as an academic audience, in relation to processes in relation to a federal government decision.

Unlike the US FARA Act, the FITS Act does not include an exemption for universities, academics, or researchers. As noted by Yee-Fui Ng and Chris Draffen (p 1124), the Commonwealth ‘deliberately chose not to include an exemption for academic or scholastic activity.’ However, Ng and Draffen (p 1126) suggest that a potential reason for not including an academic exemption may be ‘due to concerns surrounding the role of bodies such as Confucius Institutes in Australian universities’ (an account of Confucius Institutes at Australian universities is available here). Registration obligations, in theory then, could apply broadly across society and capture activity undertaken by journalists, charities, protestors, human rights organisations, and academics and researchers.

While there appears to be no immediate concern about intemperate application of the registration obligations, there should be cautious awareness of the implications of applying foreign registration requirements on civil society broadly. As Nick Robinson outlines (pp 1086-1090), under certain foreign registration schemes (most in countries that would not be considered democratic), members of civil society and human rights groups have been targeted, by way of extensive, expensive and targeted legislation that aims to stifle speech and civic engagement. In addition, these schemes often punish groups, and leaders of groups, for non-compliance, including with prison sentences. Considering these international examples, we should remain attentive to, if not immediately concerned by, broad applications of the FITS Act in Australia.

Josh Gibson is a PhD Candidate at UNSW Law and Justice.

Suggested citation: Josh Gibson, ‘Foreign influence and the implied freedom of political communication: LibertyWorks v Commonwealth’ on AUSPUBLAW (25 August 2021) <>