Comments of RMIT and ICLE Scholars to the Australian Treasury Consultation on a New Digital Competition Regime
Executive summary
The Australian Treasury’s proposed competition regime for digital platforms is flawed and should not proceed.
The policy rationale for an ex ante regime is unjustified. The Competition and Consumer Act 2010 (CCA) already provides a comprehensive framework to address concerns such as market power, unfair contract terms, and self-preferencing. The Australian Competition and Consumer Commission (ACCC) has not demonstrated any compelling reason existing competition laws are insufficient to regulate digital platforms and has not sought to enforce them against digital platforms.
The proposed regime is based on a misunderstanding of competition in the digital economy. Digital markets are characterised by dynamic competition, where innovation and technological change are the primary drivers of consumer welfare. The proposed ex ante regime, with its focus on static competition, may dampen innovation incentives and create barriers to technology diffusion, harming Australian consumers and businesses in the long run. Competition policy for digital platforms should be based on a dynamic competition approach that fosters innovation.
The proposed regulatory mechanisms are problematic. The reliance on subordinate legislation for crucial policy decisions is inappropriate, reducing parliamentary oversight. This approach lacks transparency and accountability, and may lead to unintended consequences for the digital economy.
We urge the Australian Treasury to reconsider its approach to regulating digital platforms. Instead of imposing an ex ante regime, the focus should be on enforcing existing competition laws and fostering a dynamic environment of innovation. This approach would better serve Australia’s long-term economic interests and the continued growth of the digital sector.
I. Is there need for further competition law reform?
The Proposal Paper, taking its lead from the Australian Competition and Consumer Commission (ACCC), asserts that digital platforms necessitate a new, ex ante regulatory regime in Australia.[1] This opening section of our submission will primarily focus on how the Competition and Consumer Act 2010 (CCA) already provides a comprehensive framework for addressing competition concerns in the digital economy.
The proposal for ex ante regulation is based on the premise that existing laws are insufficient to address the unique challenges posed by digital platforms. Specifically, the Proposal Paper argues that:
The characteristics and dynamic nature of digital platform markets mean that enforcement of existing economy-wide provisions of the Competition and Consumer Act 2010 (Cth) (CCA) may not on its own be sufficient to protect and promote competition, or well-suited to addressing the range and scale of competition harms identified in digital platform markets.[2]
What are these characteristics? The Proposal Paper identifies three specific issues that are already subject to regulation under the CCA.
First, market power. The existence of market power is competition agnostic. Market power typically occurs as a consequence of innovation and efficiency. A firm that innovates, whether through introducing a new product or service, improving an existing offering, implementing a novel production process, or discovering a new market, may naturally attract a larger market share.[3] However, competition law is concerned about market power in two contexts: (a) acquisitions that would result in a substantial lessening of competition; and (b) misusing market power to have the effect of substantially lessening competition in that market or another market upstream or downstream. The first context is regulated by sections 45 and 50 of the CCA – and from January 2026 the ACCC will have new administrative powers for merger control[4] following a long-running campaign. The Proposal Paper refers to the merger reforms in passing but does not consider the ACCC’s enhanced ability to regulate market power in the digital economy. The second context is regulated by section 46 of the CCA, most recently modified in 2017 following the Harper Review.[5]
Second, “Take it or Leave it” terms. Unfair contract terms are principally consumer law issues rather than competition law issues. Unfair contract terms imposed by digital platforms (or indeed any other businesses) are already addressed by the Australian Consumer Law (ACL). Part 2-3 of the ACL voids terms of unfair standard-form contracts and pecuniary penalties may be imposed for a contravention. These are general provisions and will apply to contracts with digital platforms. Given the rationale of unequal bargaining power between parties, the unfair contract provisions are appropriately limited to consumer contracts and small business contracts. Further the ACL contains other general prohibitions such as misleading or deceptive conduct[6] and unconscionable conduct[7], both of which will capture a broad range of unfair or oppressive practices against consumers and businesses.
Third, self-preferencing. Self-preferencing is a common commercial practice and an instinct for businesses – if you build something valuable, you are naturally inclined to use it to your advantage. This does not mean that self-preferencing is anti-competitive. Taking a dynamic view of competition (further discussed in section 2) incentivising innovation, driving competition on the merits, enhancing user experience and ensuring quality control (further discussed in section 4) all tend to enhance consumer welfare. However, existing provisions in the CCA are also capable of addressing concerns about digital platforms favouring their own products or services. Depending on the type of conduct, these include prohibitions on contracts, arrangements or understandings that restrict dealings or affect competition (section 45), misuse of market power (section 46), and exclusive dealing (section 47).
The “economy-wide” provisions of the Act identified above are tailored to the specific economic circumstances of each case. That is, the prohibitions on restrictive trade practices in Part IV of the CCA are with reference to competition in markets. “Market” is defined in section 4E of the CCA as a “market in Australia and, when used in relation to any goods or services, includes a market for those goods or services and other goods or services that are substitutable for, or otherwise competitive with, the first-mentioned goods or services.” This definition provides a broad degree of flexibility – encompassing digital markets – while grounding legal analysis in economic reality. The ACCC might prefer to define a market narrowly, while digital platforms facing scrutiny might favour a broader market definition. The decision of how narrowly or broad to define a market is properly one for the Court to make on the basis of economic evidence presented by the parties. The ex ante proposal proposes to skip this crucial step in competition analysis.
The Proposal Paper appears to have accepted the ACCC’s reasoning that Court proceedings are lengthy, that conduct may continue despite regulatory action, and remedies may be insufficient to address sources of harm.[8] These are general criticisms of any ex post framework, not specific to digital platforms. The ACCC has a history of consumer law enforcement actions against digital platform businesses with notable successes including against Google,[9] Meta[10], and Uber[11]. These successes prompt a crucial question: If the ACCC can enforce consumer laws against digital platforms, why has it not similarly enforced competition laws? The most recent ‘root-and-branch’ review of competition law in Australia (Harper Review) did not canvas any need for an ex ante regime or raise any significant issues with enforcing competition law against digital platforms.
The reality is that the Federal government and the ACCC have prioritised a long-running digital platforms inquiry to agitate for greater regulatory power instead of enforcement action. After nine interim reports, with a tenth report to come – and a previous iteration of the inquiry that released its own issues paper, preliminary report, and final report – there appears to be insufficient evidence of actual (c.f. hypothetical) consumer harm to bring competition enforcement action.
It would be a mistake to confuse concerns relating to enforcement with market failure. If length of time and litigation cost are the key concerns impacting competition law enforcement, then there are practical things that the Federal government might do to directly address those issues. For instance, the Treasurer (through the Statement of Expectations and Statement of Intent) might direct the ACCC to prioritise personnel and resources towards enforcement. Additionally, the Treasury might work with the Attorney-General’s Department to review Federal Court resourcing and develop law reform to expedite or streamline regulatory enforcement cases.
The Proposal Paper seeks to move to a new framework without clearly articulating the reasons why existing provisions cannot be used to enforce anti-competitive conduct in digital platforms. In doing so, the Proposal Paper overlooks the inherent flexibility and adaptability of existing competition laws, which have proven effective in addressing competition concerns and protecting consumers in various sectors, including the digital economy. The next section examines why the proposed regulatory framework reflects an outdated, industrial era understanding of competition.
II. Digital competition regulation should be based on the principle of dynamic competition rather than industrial competition
Competition policy during the 20th century sought to address industrial concentration and power that harmed consumers with high prices and limited choice. However, today, that world of industrial monopoly due to scale economies is ending. While there are still powerful economic forces pushing toward bigness – and which have produced the so-called ‘big tech’ platforms – the competitive logic that both creates and destroys these new digital giants is not the same as the previous era. Digital economies are different to industrial economies and require a different approach to competition policy. Rather than focusing on industrial concentration, the Treasury should conceptualise competition based on a dynamic model.
Digital economies have lower transaction costs than industrial economies, and scale differently. They have lower communication costs, search costs, verification costs, and networking costs, all of which increases economic complexity.[12] Business models for value appropriation and capture work differently in digital economies, often relying heavily on co- specialised assets and control of ecosystem bottlenecks. We can understand ecosystems as “groups of firms that must deal with either unique or supermodular complementarities that are nongeneric, requiring the creation of a specific structure of relationships and alignment to create value”.[13] The strategic competitive need to create these complex complementary assets and complementary technologies “are more significant than ever in a world of competing and intersecting digital platforms”.[14] In consequence, digital economies bring profound new regulatory challenges.[15]
Overall, the most important difference is in the way that competition works. Specifically, a digital economy is dominated by dynamic competition. Now of course industrial economies also experienced both static and dynamic competition, as Schumpeter long ago explained. But in digital economies, dynamic competition is the dominant force, often even in the short run. This insight should be the starting point of any new digital competition regime. This section unpacks dynamic competition and its implications.
2.1. Dynamic competition explains competition in digital markets
There are two types of competition in a modern economy and, correspondingly, two paradigms of competition economics:
- Static competition – the paradigm of neoclassical economics and modern industrial organisation theory. Static competition is price competition in a given market. It competes for existing and known rents.
- Dynamic competition – the evolutionary paradigm of complexity economics and strategy.[16] Dynamic competition is evolutionary discovery and innovation of future markets. It competes for new rents, under high uncertainty.
The central idea of dynamic competition is that there are two mechanisms by which competition unfolds. The first is static competition which works through pricing (in existing markets, with given products and technologies). The second is dynamic competition which works through innovation (i.e. the creation of new markets through innovation of new products and technologies). Dynamic competition theory acknowledges the economic logic of static competition in given markets and the consequences for consumer welfare but argues that an even larger and more important force is often at work in shaping how competition affects consumer welfare through innovation and discovery.
Dynamic competition theory has deep roots in modern economics, building on Schumpeter’s argument that innovation is the main source of consumer welfare in the long-run, and on Hayek’s observation that consumer wishes are not given information and therefore that competition is a discovery process.[17] Innovation drives competition as much as competition drives innovation. Dynamic competition comes from
the new commodity, the new technology, the new source of supply, the new type of organization – competition which commands a decisive cost or quality advantage and which strikes not at the margins or profits and the output of existing firms but at their foundations and very lives.[18]
The idea that innovation causes competition is the foundation of dynamic competition theory. It follows that a major purpose and goal of competition policy is to support and incentivise innovation.
Dynamic competition is about new entrants and incumbents engaging in new product and process development in order to create entirely new markets and product categories. In environments characterized by innovation, firms do not just look “sideways” to rivals. They look “forward” and anticipate (and create) latent competition in order to satisfy existing and future user/customer needs, thereby unlocking potential demand and stimulating economic development and growth. Frequent new product introductions, often followed by price declines, are commonplace.[19]
The mechanism by which dynamic competition works is innovation to create future markets through the strategic development of dynamic capabilities.[20] These are capabilities to innovate, and involve organisational, technological, and managerial capabilities.
We define dynamic capabilities as the firm’s processes that use resources – specifically the processes to integrate, reconfigure, gain and release resources – to match and even create market change. Dynamic capabilities thus are the organisational and strategic routines by which firms achieve new resource configurations as markets emerge, collide, split, evolve and die.[21]
Dynamic capabilities are the sensing, seizing, transforming skills that enable a firm to identify, develop, market, and sell innovative products and are a form of evolutionary capital. The strategic management literature finds considerable evidence that strong dynamic capabilities are the leading cause of success in big tech firms.[22] The ability to build new internal and external capabilities for innovation enables firms to develop competitive advantage and thus enhances competition in advanced markets. These decisive capabilities for dynamic competition must be built through investment, including acquisition. Firms differentially develop these strategic capabilities, which can be difficult to observe, and often are only evident ex post.
Modern competition policy tends to focus on static competition. There are several reasons for this. The first reason is that the theory supporting it is both more widely known (standard industrial organisation) and operationalizable (e.g. by observation of firm concentration measures and prices). On the other hand, dynamic competition economics is less well- known (it is usually taught as advanced economic and business strategy training rather than as undergraduate introductory courses) and requires focus on more complex and difficult to measure objects (innovation inputs, entrepreneurial processes, capabilities, ecosystems).
Another reason competition policy analysis and enforcement rarely considers innovation- focused dynamic competition is because it is harder to make arguments and assemble evidence. Because dynamic competition is future oriented evidence of harms (and benefits) does not exist ex ante but must be constructed or inferred. That requires specialist capabilities that regulators generally lack. Consequently, it is easier to base policy and enforcements on static economic arguments about how competition works to infer consumer benefits from existing (rather than future) consumers.[23] The upshot is that a competition policy that focuses on dynamic competition is harder to do. Even still, it can be done and is worth doing – and especially so in the context of digital competition.
2.2. Digital competition policy should be primarily based on dynamic competition
Modern competition policy is centred on the paradigm of static competition that often struggles to capture dynamic capabilities. This can lead to a misjudgement of the very market processes needed to build these capabilities, such as mergers and acquisitions, which are essential for firms to compete effectively. While the focus of static competition may be less problematic in traditional industrial markets, it becomes a significant issue in the context of the digital economy where innovation and dynamic competition are the primary drivers of consumer welfare.
The proposed new digital competition regime risks perpetuating this oversight. A dynamic competition approach should seek to understand how to bring up the long term faster, and to safeguard and support the short term capabilities that enable dynamic competition.
The standard theory of digital economics builds on the significance of non-rivalry, the business model of platforms (and zero prices on one side of the market) and the importance of information as an economic good and as central to value appropriation.[24] The static IO model of platform competition then emphasises the monopoly power (winner-take-most effects) that large platforms have due to network effects and consumer switching costs.
There is no innovation aspect of this model.
In contrast, a dynamic competition perspective recognises that firms are not merely competing for existing rents but are striving to discover and capture future rents through investments in innovation. This requires building competencies and capabilities for innovation that may be difficult for regulators to distinguish from monopolistic behaviour ex ante. The capabilities that matter to dynamic competition are managerial, technological, and organisational. The efforts that firms go to by building capabilities to discover and capture these rents will be difficult to distinguish for competition regulators from monopolistic behaviour. It should be noted of course that high profit does not necessarily mean monopoly rents from exploitation of monopoly power. Rents may come from superior efficiency, including dynamic efficiency or innovation. This is why ex ante rules designed to mitigate static competition abuses will inadvertently harm the dynamic competition mechanisms that ultimate benefit consumers.
Fostering dynamic competition in the digital economy requires a shift in focus. It also requires a policy that rewards the search for excellence under uncertainty, and that displays higher tolerance towards positions of structural monopoly that arise from genuine innovation.
2.3. Digital markets bundle platform with product
Consider one further observation about why digital economy competition is fundamentally different – namely that digital dynamic competition involves innovation not only at the product level (i.e. of the good or service), but also in the institutional layer below on which that service is operationalised and delivered.
In other words, in digital industries ‘large tech firms’ often produce both a product plus the network on which it runs. They are not factories. They are usually platforms. This is apparent in old digital industries, such as telecommunications and social media, as well as new decentralised computational industries such as blockchain and artificial intelligence. Bitcoin, for instance, is not just a digital token (i.e. a money), but a token plus the transfer, settlement, and security layer on which it operates. The key point is that consumer benefit accrues not just to the token (and its price) but to the properties of the network.
Competition policy should therefore be just as interested in the innovation potential and capabilities of the platform or network as the good or service itself. Indeed, from the long-run consumer perspective, innovation on the network or platform is generally the precondition for innovation in the good or service. A major source of innovations that build future markets are dynamic capabilities assembled in modern business ecosystems directed at platform innovation.[25] Unfortunately, the proposed ex ante framework fails to adequately address these considerations and, as discussed in the next section, raises significant concerns.
III. Concerns with the design of the legislative framework
The proposed regulatory framework consists of broad obligations and service-specific obligations that are applied to all services (digital platforms in respect to the specific services that they provide consumers) that have been designated by the relevant minister (presumably the Treasurer or Assistant Treasurer). Broad obligations are general rules that all designated entities must follow and are proposed to be implemented in primary legislation. They are designed to prevent anti-competitive behaviour, promote transparency, and ensure fairness. Examples include prohibitions on anti-competitive self-preferencing, restrictions on tying, measures to prevent impediments to consumer switching, and rules ensuring interoperability and transparency. They are highly general and in part duplicative of existing competition law, as discussed above.
Service-specific obligations apply the broad obligations to the direct context of the service. It is proposed that these are tailored rules for different types of services (e.g., app marketplaces, ad tech). They specify how designated entities must comply with broad obligations. These obligations are implemented through subordinate legislation.
All designated entities must comply with broad obligations regardless of whether service- specific obligations exist. If service-specific obligations are established, failing to meet them constitutes a breach of the corresponding broad obligation.
The proposed framework is extremely ambitious in its design. It seeks to design an ex ante regulatory system that can regulate not just a specific set of digital platforms with known and specific business models but all digital platforms with heterogeneous business models in a rapidly shifting market environment, as well as all future digital platforms that might operate in Australia. This presents a challenge for ex ante regulation that ex post common law regulatory systems are better equipped to address. This section focuses on our concerns with the structure and design of the proposed legislative framework, moving beyond the rationale and theoretical underpinnings.
3.1. Subordinate legislation is inappropriate for substantive policy
The structure outlined in the Proposal Paper attempts to establish broad principles of conduct in digital markets above and beyond the existing competition law while recognising that digital markets are heterogeneous and what it means to be (for example) ‘transparent’ or ‘interoperable’ in any given digital market is highly contextual. We have raised concerns above with the need for new competition law along these lines. Here we consider a separate concern with the proposed framework: introducing the service specific obligations as subordinate legislation is inappropriate. If the Proposal Paper is any guide, the scale and scope of service-specific obligations are of such significance that they should be enacted through primary legislation.
Consider for example the question of interoperability. The Proposal Paper argues that “Restrictions on interoperability that limit effective competition” would be appropriate for broad obligations.[26] Interoperability might seem to be prima facie desirable. But interoperability is a highly complex, multifaceted concept that has significantly different meanings and different modes of operation. As one of us has argued in the Internet Policy Review, “the study of interoperability per se is highly domain-specific and used in many fields to describe a variety of system characteristics”.[27] Interoperability can consist of technical interoperability (the ability of different technological systems, platforms, or devices to communicate and exchange information), syntactic and semantic interoperability (how data is structured for exchange, and how data is understood), and organisational interoperability (and capable organisations are to actual share resources and align processes). Developing specific criteria on the processes by which heterogeneous services might be considered to be ‘interoperable’ is not a technical matter that can be delegated to regulators.
The Proposal Paper’s examples illustrate this point. The ACCC has argued that competition is being limited by, for example, “mobile OS providers not providing third-party providers of apps and services with reasonable and equivalent access to hardware, software, and functionality”.[28] It might be reasonable for parliament to delegate to regulators and subordinate legislation what is meant by “reasonable and equivalent” or what appropriate level of hardware access, software interface or functionality would be considered to be truly interoperable (noting, of course, that these are complex decisions with deep policy impacts and regulatory costs). But it is not reasonable to delegate the entire question of whether interoperability in this form is a minor question of implementation. There are an enormous variety of mechanisms by which the principle of ‘interoperability’ could be regulated, and these should be understood as policy rather than implementation questions.
This is true for each of the domains in the Proposal Paper. As we discuss below, rules governing self-preferencing, in-app payment restrictions, and access to APIs and system- level features for app stores have far-reaching consequences, not just for the affected platforms but for thousands of Australian businesses and millions of consumers. Similarly, ad tech services, which form the backbone of digital advertising, are governed by complex supply chains. Obligations to ensure transparency in auction processes, data access, and competition between ad networks affect not only the business models of dominant platforms but also the viability of media organisations, independent publishers, and advertisers across the economy. Given the economic weight of these sectors and their central role in digital commerce, decisions about their regulation are substantive matters of policy.
Moreover, the implications of these obligations extend well beyond market efficiency and competition law—they touch on fundamental questions of consumer rights, data governance, and national digital sovereignty. Interoperability is not just a technical specification but a policy choice that determines the degree to which different platforms and services can function together, shaping user control, market entry for new firms, and long-term technological innovation. Similarly, rules governing algorithmic transparency in search and recommendation engines influence the visibility of businesses and the diversity of content Australians engage with daily. If Australia follows suit in regulating these industries, the legal framework must reflect the scale of intervention.
3.2. Service-specific obligations should not be subject to a lower standard of public scrutiny and democratic debate through the subordinate legislation mechanism.
Each ‘service-specific’ obligation as described in the Proposal Paper represents the wholesale regulation of major and economically critical industries and will affect every Australian consumer. Subordinate legislation is however subject to a much lower bar of public debate and scrutiny. Handing off significant policy areas to subordinate legislation reduces the democratic discussion necessary to provide those policies with parliamentary legitimacy.
It is correct that subordinate legislation is subject to disallowance by the parliament. For minor areas of implementation and regulatory specifics, the disallowance process works well, sometimes surfacing controversial policy decisions that parliament disagrees with the executive on. But at the same time the disallowance mechanism is designed as an emergency democratic backstop – a handbrake on the executive to ensure parliamentary supremacy. Parliament faces an “avalanche” of subordinate legislation to which it can only draw so much attention.[29] By contrast, primary legislation undergoes a comprehensive process involving multiple readings, debates, and committee reviews in both legislative chambers, ensuring thorough scrutiny and democratic accountability. This rigorous procedure allows elected representatives to examine, amend, and vote on proposed laws, reflecting the will of the electorate. Primary legislation must be positively approved by the parliament – it is not the subject of a structural assumption that it will be implemented unless parliament objects.
The disallowance mechanism also only functions effectively as a democratic protection while the opposition and crossbench are able get a majority for disallowance. While it has been nearly two decades since a government controlled the Senate, the desirability of this legislative framework should not be predicated on the government being a minority in the upper house.
3.3. The process of subordinate legislation reduces the scrutiny of regulation from other affected government departments, particularly departments focused on cybersecurity.
The parliamentary process does not only afford parliamentarians and the public to scrutinise legislation – it also opens opportunities for other government departments to scrutinise proposals. This is particularly important in the case of digital market regulation. Below we describe some of the critical national security issues that the specific proposals suggested by the Proposal Paper might trigger. Regulations under competition policy intersect with cybersecurity concerns as the technical structure of software which has access to important or private information about users is vulnerable to bad actors, both state and non-state.
Recent controversies around the social media application TikTok and the AI application DeepSeek – both of which have been banned for use on government devices – underline the point that policy decisions to force changes to technical systems are not simply the domain of competition regulation, and not simply the provenance of Treasury or other economic ministries. Where subordinate legislation is not subject to full cabinet consideration and discussion, this means that relevant ministries and departments (particularly national security agencies) will have less visibility over critical regulations that affect the cybersecurity of Australian citizens.
3.4. It is standard practice that matters of policy significance should be in primary legislation.
That matters of policy significance should be implemented through primary legislation is a settled matter of Commonwealth parliamentary practice. The 2017 Legislation Handbook outlines the principles by which rules should be considered as primary rather than subordinate legislation (section 1.10). We draw attention to the following principles, each of which describe policies which are proposed to be ‘service-specific’ obligations:
(b) significant questions of policy including significant new policy or fundamental changes to existing policy;
(c) rules which have a significant impact on human rights and personal liberties;
(d) provisions imposing obligations on individuals or organisations to undertake certain activities (e.g. to provide information or submit documentation, noting that the detail of the information or documentation required may be included in subordinate legislation) or desist from activities (e.g. to prohibit an activity and impose penalties or sanctions for engaging in an activity);
(j) procedural matters that go to the essence of the legislative scheme;
In addition, section 5.66 of the Legislation Handbook notes that while using subordinate legislation to simplify primary legislation is desirable, this should not come at the expense of reducing parliamentary control over government policy.
This is reflected further in the Senate Scrutiny of Delegated Legislation Committee’s guidelines:
Significant elements of a program of national significance or a regulatory scheme should ordinarily be included in primary rather than delegated legislation, due to the higher level of parliamentary scrutiny associated with the legislative process for primary legislation.
3.5. The designation process leaves too much discretion to the minister.
The proposed designation process grants the minister significant discretion in determining which digital platforms are subject to the new regulatory framework. While the ACCC is tasked with conducting investigations and making recommendations, the final designation decision rests solely with the minister (and as a matter of administrative law should not merely follow the advice). This structure grants excessive ministerial authority, as it enables political influence over which platforms are regulated and under what conditions.
Furthermore, while the ACCC is responsible for gathering information and consulting stakeholders, there is no clear requirement for public disclosure of its findings before the minister makes a designation decision, reducing transparency and accountability.
Additionally, while the designation process is proposed to rely on a mix of quantitative thresholds and qualitative assessments, it is the minister who ultimately determines how these factors are weighed. The proposed approach allows the minister to direct the ACCC to conduct investigations at their discretion and make designation decisions with limited parliamentary oversight. This process risks consolidating too much regulatory power in the hands of a single government official.
The five-year designation period, while apparently intended to balance regulatory stability with regular review, also means that platforms could be subject to significant obligations that are inappropriate and harmful to both competition and innovation. The recent evolution of the market for search – which is increasingly being challenged and modified by artificial intelligent systems both within ‘traditional’ search platforms (like Google’s Gemini search function at the top of the Google search results) and outside them (like the search functions built into Anthropic’s Claude, OpenAI’s ChatGPT, and aggregators such as Perplexity) shows how rapidly digital markets can change.
3.6. All obligations both broad and service-specific should be subject to parliamentary scrutiny though primary legislation.
Australia should not adopt an ex ante competition policy framework. However, if it does then all obligations should be written into primary legislation so that they are subject to proper parliamentary scrutiny. The experience from ex ante regimes in other jurisdictions demonstrate that they are not mere technical changes – they are radical market interventions requiring substantive changes to how major digital platforms operate.[30] The current proposed two-tier framework, by relegating “service-specific” obligations to subordinate legislation, fails to acknowledge that these decisions will reshape Australian business models, innovation incentives, and cybersecurity risks. We should not rely on disallowance processes to ensure we grow Australia’s prosperity and maintain our safety.
The claim that subordinate legislation provides necessary flexibility and adaptability ignores the profound market consequences at stake. Consider the real impacts already seen overseas, such as a 17.6% drop in hotel booking clicks for one company following the introduction of ex ante competition rules.[31] In Australia, these changes would affect how local businesses advertise and how consumers compare travel options and make purchases. While technology may evolve rapidly, the fundamental reshaping of business models and market structures demands thorough legislative scrutiny, not just hopes of disallowances.
Australia needs parliamentary scrutiny and debate over obligations that grapple with:
- Trade-offs between competition and national security: Opening platforms to third-party access creates security vulnerabilities that jeopardise Australian user data. Mandatory third-party app stores significantly complicate security scanning and malware prevention, leading digital platforms to combat the impacts of these changes through imperfect safeguards.[32]
- Escalating implementation and compliance burdens: Building new API infrastructures, security monitoring systems, and support systems for third parties is not merely technical Estimates of the compliance cost burden of the DMA on US firms has been in the tens of billions of dollars.[33] Whatever the cost burden in Australia, it will ultimately fall on Australian people and businesses who rely on these critical services.
- Weaponisation by small competitors to impose costs on large competitors: Rather than investing resources into developing their own capabilities, smaller firms could use these service-specific obligations to transfer costs onto larger rivals. For instance, instead of building interoperability with existing hardware, firms might demand access under regulation. This strategic move would effectively shift R&D costs to competitors, creating perverse incentives that reward regulatory rent seeking over genuine innovation. There is evidence from the EU DMA experience that smaller “middleman” firms are continuing to push for more changes to the major digital platforms, pushing further costs onto them through regulation.[34] The EU DMA experience also shows app stores being forced into approving apps that may go against community values and expectations.[35]
The issues outlined in this section highlight the need for a more transparent legislative framework that prioritises policy scrutiny. The inappropriate reliance on subordinate legislation and ministerial discretion raises concerns about the potential for regulatory overreach and unintended consequences. In the next section, we will examine specific examples of service-specific regulations and their potential economic impacts, further illustrating the need for a more cautious and considered approach to regulating digital markets.
IV. The example service-specific regulations could have significant and harmful economic consequences
These immediate impacts point to deeper and more troubling consequences for Australia’s future. When regulation increases costs and complexity while enabling rent-seeking behaviour, it alters innovation incentives. There are two main ways that this manifests. First, major platforms’ willingness to invest in innovations that ultimately benefit Australian consumers and businesses. Second, new barriers to the adoption and diffusion of technologies into the Australian economy. Both are major challenges given that technological change is the primary driver of economic growth and prosperity. In this section we will consider both consequences in detail.
4.1. Many obligations will directly dampen incentives to invest in technology and innovation.
The Proposal Paper’s suggested requirement that a digital platform must “provide third-party providers with reasonable and equivalent access to hardware, software, and functionality”[36] will lead designated companies to factor mandatory sharing into every major R&D decision. Consider a tech company contemplating a large investment in developing new AI security features. When making the decision to make this risky investment a proposed ex ante regime would threaten that, if successful, any breakthrough might need to be immediately shared with competitors. How would they recoup R&D costs or build a competitive advantage? When combined with the security and compliance burdens outlined above, this creates a clear disincentive for investing in innovations that would benefit Australian users.
4.2. Proposed obligations also raise barriers to the diffusion of technologies – preventing innovations from reaching Australian consumers and businesses.
Technologies and products don’t only need to be invented and developed, they also need to spread. Responding to various ex ante obligations, in other markets we have seen market distortions that hurt domestic consumers and businesses by requiring major changes by digital platforms. For instance, Google has undergone a series of changes in how maps and hotel bookings interact[37] as well as the reduced visibility of the Google flights feature.[38] Meta delayed the rollout of its then-new social media platform, Threads, because of regulatory uncertainty[39] while Apple withheld the launch of frontier AI features because of potential security risks.[40]
Australia is a small economy and should expect that similar ex ante rules will also deter major firms from operating here. Not only will digital platforms likely reduce R&D investment knowing they cannot protect their innovations, but they might also delay or downgrade product releases rather than pay the costs of adapting to radical Australian requirements.
Below we outline some examples of service-specific obligations and their major economic consequences:
| Service-specific conduct to be addressed in subordinate legislation | Major economic impacts for Australian businesses and consumers |
| “App marketplaces providing more favourable treatment to their own apps in app store search result rankings” |
|
| “Mobile OS providers not providing third-party providers of apps and services with reasonable and equivalent access to hardware, software, and functionality” |
|
| “App marketplaces restricting developers’ ability to communicate to consumers regarding alternative payment or purchase channels” |
|
These are just some examples of specific conduct under just two types of digital platforms into the proposal. Even from these examples it is clear even from these examples that the artificial distinction between “broad” and “service-specific” obligations threatens to push major economic interventions into subordinate legislation. We need proper democratic oversight of changes that could affect billions in investment, reshape fundamental business models, and create significant security risks. All obligations – including those characterised as “service-specific” – should be subject to parliamentary scrutiny through primary legislation.
This is also coming at a time when these firms are being revolutionised by new AI platforms. The traditional digital platforms targeted by this regulation are themselves navigating the integration of AI capabilities throughout their products and services. Imposing rigid structural obligations now, without fully understanding how AI will reshape these markets, risks constraining innovation just as our digital economy undergoes fundamental change.
We are also yet to see how general obligations might apply to emerging platforms such as AI models themselves. How would data portability requirements apply to large language models where user interactions help refine and improve the model? Would users have a right to port their chat histories and custom instructions between AI services? Large language models increasingly serve as search interfaces, but with fundamentally different architectures than traditional search engines. How would self-preferencing obligations apply here? If an AI platform promotes its own products or services in responses, would this constitute unfair self-preferencing? The proposal’s focus on traditional digital advertising may not adequately capture these new forms of commercial influence. Even more fundamentally, many AI platforms operate as both infrastructure providers and application developers. They offer APIs for others to build on while developing their own consumer-facing products. Anti- tying and interoperability requirements could significantly impact this dual role. Would AI platforms be required to offer competitors equal access to their most advanced models?
Could they bundle their models with implementation tools and services? These questions go to the heart of competition in AI markets and further underscore the need for deep parliamentary scrutiny of all obligations.
Conclusion
The proposed ex ante competition regime for digital platforms raises significant concerns. The Australian Treasury should reconsider its approach – focusing on enforcing existing competition laws, policy based on dynamic competition, and incentivising innovation to benefit Australian consumers over the long-run.
[1] Treasury, ‘Digital platforms – a proposed new digital competition regime’ (Proposal Paper, December 2024) (Proposal Paper); Australian Competition and Consumer Commission, ‘Digital Platform Services Inquiry – Discussion Paper for Interim Report No. 5: Updating competition and consumer law for digital platform services’ (Discussion Paper, February 2022) (ACCC Digital Platforms Inquiry Interim Report 2022).
[2] Proposal Paper, p. 4.
[3] Schumpeter (1934)
[4] Treasury Laws Amendment (Mergers and Acquisitions Reform) Act 2024 (Cth).
[5] Competition and Consumer Amendment (Misuse of Market Power) Act 2017 (Cth).
[6] Section 18 ACL.
[7] Sections 20-22 ACL.
[8] See: ACCC Digital Platforms Inquiry Interim Report 2022, p. 51.
[9] ACCC v Google LLC (No. 4) [2022] FCA 942; ACCC v Google LLC (No. 2) [2021] FCA 367.
[10] ACCC v Meta Platforms Inc [2023] FCA 842.
[11] ACCC v Uber BV [2022] FCA 1466.
[12] Goldfarb and Tucker (2019)
[13] Jacobides et al (2016: 2263)
[14] Teece (2018: 1382)
[15] Jacobides et al (2024)
[16] Teece (2023)
[17] Schumpeter (1943); Hayek (1945).
[18] Schumpeter (1943: 85)
[19] Teece (2023)
[20] Teece et al (1997), Pisano and Teece (2007), Sutton (2012), Teece (2007, 2018, 2019)
[21] Eisenhardt and Martin (2000: 1107)
[22] Helfat and Peteraf (2015)
[23] Potts (2023)
[24] Goldfarb and Tucker (2019)
[25] Teece (2012), Teece (2018), Petit and Teece (2021).
[26] Proposal Paper, p. 19.
[27] Berg (2024).
[28] Proposal Paper, p. 22.
[29] Kirrily Schwarz (2020) “Who is making our laws? The separation of powers in 2020” LSJ. 11 November, https://lsj.com.au/articles/who-is-making-our-laws-the-separation-of-powers-in-2020/
[30] On some of the ways “gatekeepers” under the EU’s DMA have responded see: https://www.theverge.com/2024/3/6/24091592/eu-dma-competition-compliance-deadline-big-tech- policy-changes
[31] See: https://truthonthemarket.com/2024/03/12/the-broken-promises-of-europes-digital-regulation/.
[32] See: https://www.apple.com/newsroom/2024/01/apple-announces-changes-to-ios-safari-and-the- app-store-in-the-european-union/; and see further: https://developer.apple.com/security/complying- with-the-dma.pdf
[33] See Kati Suominen (2022) “Implications of the European Union’s Digital Regulations on U.S. and EU Economic and Strategic Interests”, Report for the Centre for Strategic Studies. available online, https://csis-website-prod.s3.amazonaws.com/s3fs-public/2023-02/221122_EU_DigitalRegulations- 3.pdf?VersionId=04r7zBzS2kHNhsISAqn4NkC6lGNgip7S
[34] See: https://medium.com/chamber-of-progress/the-digital-markets-acts-statler-waldorf-problem- 2c9b6786bb55
[35] See: https://9to5mac.com/2025/02/03/apple-forced-to-approve-porn-app-on-eu-iphones-due-to- dma/
[36] Proposal Paper, p. 22.
[37] See: https://blog.google/around-the-globe/google-europe/an-update-on-our-preparations-for-the- dma/ https://blog.google/around-the-globe/google-europe/dma-compliance-update/ https://blog.google/around-the-globe/google-europe/complying-with-the-digital-markets-act/
[38] See: https://blog.google/around-the-globe/google-europe/an-update-on-our-preparations-for-the- dma/
[39] See: https://www.theverge.com/23789754/threads-meta-twitter-eu-dma-digital-markets
[40] See: https://www.cnbc.com/2024/06/21/apple-ai-europe-dma-macos.html