08/29/2024 | Press release | Distributed by Public on 08/30/2024 08:57
Contents
International Regulatory Developments 2
Major Developments in Digital Platform Markets 3
Emerging Issues: Cloud Computing and AI 4
The Australian Competition and Consumer Commission's (ACCC) Digital Platforms Branch's five-year Digital Platform Services Inquiry (2020-2025) is a comprehensive project focusing on whether Australia's existing legal framework, including its competition and consumer protection laws, is adequate to address the issues raised by the digital economy. The Digital Platform Services Inquiry - March 2025 - Final Report Issues Paper published on July 25, 2024 (Issues Paper) focuses on international regulatory developments, major developments in digital platform markets (highlighting online private messaging and app marketplaces), and potential or emerging issues (such as online gaming, small business issues in cloud computing, and generative artificial intelligence (AI)) as three areas of interest for its upcoming final report and for which it is interested in further stakeholder engagement.[1]
The Information Technology and Innovation Foundation (ITIF) appreciates the opportunity to provide comments on the Issues Paper from the standpoint of promoting innovation in Australia. ITIF is a nonprofit, non-partisan public policy think tank based in Washington, D.C., committed to articulating and advancing pro-productivity, pro-innovation, and pro-technology public policy agendas around the world that spur growth, prosperity, and progress. ITIF's comment proceeds in five parts.
First, the international landscape varies considerably with respect to digital regulation. While the European Union (EU) has implemented regulations like the Digital Markets Act (DMA) and Digital Services Act (DSA), this does not represent a uniform, and certainly not an optimal, trend that countries like Australia should follow. Second, regulation in the digital sector should be a response to market failures, which do not appear to exist in Australia's online private messaging and app marketplace sectors. Third, the Issues Paper emphasizes potential competition concerns in cloud computing and Generative AI markets, which are belied by the dynamic and growing nature of these sectors, as well as the natural and needed economies of scale. Fourth, ITIF recommends that the ACCC consider digital models beyond those mentioned in the Issues Paper, such as those of the United States, Singapore, and Taiwan, which do not involve the heavy-handed digital regulation that can stifle the very innovation Australia seeks to foster. A brief conclusion follows.
There is no uniform international trend toward digital regulation. To be sure, there has been a tendency among some countries to implement specific regulations designed to address what are perceived to be the unique characteristics and competitive shortcomings associated with digital markets. For example, the EU's Digital Markets Act (DMA) and Digital Services Act (DSA) are paradigmatic examples of a regulatory response to the perceived problems with digital markets.[2]However, other jurisdictions have taken very different policy approaches. Indeed, consistent with the continuously evolving, dynamic, and growing nature of digital markets, some jurisdictions like the United States, Taiwan, and Singapore have deliberately avoided going down the EU's path of digital regulation.[3]
Recent history provides a good reason why countries may be hesitant to embrace regulation. Beginning in the late 1970s, the United States embarked on a significant deregulatory program across various industries, such as telecommunications, railroads, and airlines, which had previously been heavily regulated. This shift was driven by a bipartisan desire to promote economic welfare and growth, reduce administrative burdens and regulatory barriers to entry, and avoid problems associated with the government picking winners and losers.[4]While the results varied by industry, deregulation worked as a general matter. For example, deregulation of the media industry led to a flourishing of innovation and competition by allowing the previously regulated cable competitors to challenge the broadcast incumbents and helped create the diverse and dynamic media landscape the United States enjoys today.
Moreover, even for those countries that have adopted digital regulation, many have pursued models very different from the DMA. Indeed, and in contrast with any one-size-fits-all approach, regulatory regimes differ based on criteria like the scope of their application (i.e., whether they target specific companies or apply industry-wide) and whether they allow for procompetitive justifications. For example, whereas the DMA targets predominantly American companies, Brazilian and Indian digital regulation is far broader, and encompasses smaller domestic firms within their regulatory requirements. Similarly, and again unlike the DMA, the digital regimes in the United Kingdom and South Korea all eschew per se rules and instead allow covered firms to have an opportunity to offer procompetitive justifications for at least some of their behavior.
At bottom, while the ACCC's Issues Paper highlights several key jurisdictions implementing digital market regulation, it must not commit a selection bias error-namely, failing to recognize that key jurisdictions like the United States, Singapore, and Taiwan have opted not to implement digital competition regulation and in so doing heed the lessons of recent economic history about the pitfalls of regulation, which are especially pertinent in dynamic markets. Moreover, the ACCC should not overlook that, in many respects, the DMA is one of the worst possible regulatory models, and in particular by virtue of its targeting of American firms and failing to allow gatekeepers to offer procompetitive justifications for their conduct.
The Issues Paper specifically highlights its reports on Online Private Messaging and App Marketplaces as areas where it intends to provide updated information in its final report. However, it is critical that in its final report the ACCC not miss the proverbial forest for the trees: regulatory intervention in these and any other digital sectors should be a response to a genuine market failure. With respect to online private messaging, the Issues Paper correctly notes that this is a sector that has seen "significant growth"-a finding hardly consistent with a failing market.[5]Moreover, not only are many online messaging services free to users, which again belies the idea of market failure, but innovation continues to flourish. For example, just recently, Apple introduced its Intelligence offering, which includes a sophisticated AI text tool.[6]
In the online private messaging sector, there is no evidence of consumer harm resulting from a lack of competition. This conclusion is supported not only by the aforementioned market performance metrics but also by its overall structure. Indeed, contrary to the 2020 Online Private Messaging Report findings, the market demonstrates more than sufficient competition that enhances consumer welfare in the form of a variety of options to Meta's Messenger and WhatsApp services. Specifically, not only do Apple and Google offer alternative private messaging products that are ubiquitously used by consumers, but over-the-top providers like Snapchat, Signal, Telegram, and others also offer widely used private messaging services-for free. Additionally, messaging is offered as part of a suite of other commonly used platforms like X, Linkedin, TikTok, Zoom, Teams, and many more, which also provide users with a wide variety of options. At bottom, this plethora of alternatives ensures that consumers are not restricted to a single service and mitigates concerns about consumer harm: in a market characterized by such a variety of choices, the likelihood of consumer harm is minimal. And, importantly, regulation should be driven by evidence of negative effects on consumers rather than the mere presence or absence of a certain number of competitors.
The Issues Paper also notes concerns about barriers to entry, and specifically the existence of "identity-based network effects" that provide "significant competitive advantages over smaller suppliers of standalone services in Australia."[7]But the existence of network effects in no way guarantees a winner-take-all market structure. Specifically, it has long been recognized how many platform markets are characterized by multi-homing strategies where firms prefer using multiple alternatives, which increases competition in the market.[8]Indeed, the existence of the numerous messaging options discussed above is consistent with consumer multi-homing strategies in the online private messaging space, which underscores the existence of a competitive market dominated by no single platform where consumers and developers can choose the product that best suits their individual preferences.
The app marketplace tells a similar story. According to the ACCC's report, "Apple and Google's dominance in mobile OS, combined with the control exerted over the app marketplaces permitted into their mobile ecosystems, means that the App Store and the Play Store control the key gateways through which app developers can access consumers on mobile devices."[9]However, not only does the existence of two large firms that benefit from network effects not at all imply market failure, but the argument that Apple and Google's strength in mobile operating systems implies app marketplace dominance is not reflected by reality. Indeed, there are a number of app store alternatives, such as Amazon Appstore, Samsung Galaxy Store, and more, which actively compete to attract both app developers and consumers.
In addition to this structural diversity, performance metrics like price, output, and innovation provide critical data points on whether the app marketplace is working. And, as commentators have estimated, Australia's application market grew $1 billion in four years to $2.6 billion in 2023-an increase of almost two-thirds or over 15% on average per year.[10]Moreover, the app marketplace sector remains dynamic and innovative. For example, firms like Apple not only continue to invest huge amounts in quality and security with their iOS upgrades, but continue to offer new App Store features like In-App Events, which allows developers to promote app related events to users and has been reported as leading to a substantial increase in app revenues.[11]
The Issues Paper identifies several "potential or emerging competition and consumer issues" previously unexamined by the ACCC but which nonetheless may be relevant to the Final Report.[12]One of these is described in terms of "potential competition and small business issues" in cloud computing, with the Issues Paper referencing the ACCC's Report on expanding digital platform ecosystems, which itself highlights concerns about potential anticompetitive conduct by firms with substantial market power.[13]However, it is important to clarify that "substantial market power" is often a synonym for "monopoly power," which is typically shown by a firm having a market share greater than 66% that is protected by barriers to entry.[14]
So understood, there are no firms with "substantial market" or "monopoly" power in the cloud space. While the market is characterized by high fixed costs, economies of scale, and innovation competition-and thus, like other infrastructure markets, not conducive to a more decentralized market structure-Amazon, Google, Microsoft, and several other large firms vigorously compete on both price and innovation. Indeed, this dynamic is consistent with the inverted-U relationship that defines many scale and innovative intensive industries where scale economies drive innovation and productivity.[15]And, importantly, there is as of now no evidence of anticompetitive conduct that would justify regulation.
The ACCC specifically raises concerns that exclusionary practices like bundling, tying, or self-preferencing allow cloud providers to leverage their supposed market power and gain advantages in other markets. But this type of conditioning behavior, whether contractual (tying), price-based (bundling), or by design (technological tying and self-preferencing), does not only not usually result in anticompetitive harms but often admits to strong procompetitive justifications that benefit consumers-even when done by large firms. Indeed, as has long been understood, tying, bundling, and technological integrations like self-preferencing-which is ubiquitous throughout digital markets-not only play an important role in preventing inefficiencies associated with free riding, but can benefit consumers through increased convenience and introducing them to new offerings, better value, and a more integrated product ecosystem.
The Issues Paper also identifies several broad issues in the area of generative AI. But, here again, concerns about market failure and pervasive anticompetitive conduct are untimely. First, as the Issues Paper concedes, "this technology continues to expand and develop at a rapid pace," which suggests regulation is at least premature.[16]And, while it is true that there are a number of reasons why (as with cloud) scale is critical in the AI space, firms like OpenAI, Anthropic, Mistral, Cohere, and others are a testament to healthy Schumpeterian competition and specifically startups challenging the established digital players. Indeed, this dynamism confirms, as ITIF has elsewhere explained, "there is no evidence of significant entry barriers" and specifically that "concerns about data being an entry barrier in AI are speculative and unsubstantiated."[17]For example, new firms like OpenAI and Anthropic have developed market-leading models without having the amount of extensive user data enjoyed by large platforms.[18]
With AI, the ACCC also raises fears about "anticompetitive self-preferencing, tying, and data access restrictions."[19]While it is true that such behavior can harm consumers in some cases-for example, self-preferencing by a monopolist to exclude competitors and harm consumers-it's important to recognize that not all self-preferencing is inherently problematic. That is, there are several situations where self-preferencing can be procompetitive and beneficial. Additionally, while there can also be competition concerns about refusals to share data, it is also crucial to consider that such refusals can preserve the strong innovation incentives that might otherwise be weakened by compulsory access schemes. Moreover, as ITIF has previously highlighted, as a general matter, cognizable exclusionary conduct harms stemming from the existence of vertically integrated AI firms "have not materialized."[20]And, regarding concerns about algorithmic collusion, leading U.S. antitrust officials have made clear that existing antitrust principles are already well equipped to deal with these issues to the extent they arise and specifically that sophisticated algorithms do not by themselves "create novel liability scenarios."[21]
For these reasons, ITIF offers the following recommendations to the ACCC:
▪Digital market regulation is not an international best practice: Global approaches to regulating digital markets are not uniform. While the ACCC highlights the EU's heavy-handed DMA, it is important to recognize that major jurisdictions such as the United States, Singapore, and Taiwan have chosen not to implement similar comprehensive digital market regulations. The United States, for instance, has some sector-specific laws, but it has not opted for an overarching digital market law-and for the better, as this helps maintain a more innovation-friendly environment. Furthermore, a wide range of approaches are being considered among the countries that have adopted some form of digital regulation, reflecting diverse policy objectives and regulatory philosophies.
▪Digital competition policy should be evidence based: Regulation in the digital sector should be a response to clear evidence of market failure and pervasive anticompetitive conduct that the existing competition policy enforcement regime is unable to sufficiently police. In the case of the online private messaging and app marketplace sectors highlighted in the Issues Paper, market performance appears to remain robust, with no clear signs of failure. And, rather than improve upon the status quo, introducing regulation in these markets could hinder innovation.
▪Let cloud computing and generative AI markets continue to flourish: The Issues Paper's concerns about market dominance in cloud computing and generative AI are not well substantiated. On the contrary, while characterized by a need for scale to drive innovation, these sectors remain competitive, with Schumpeterian entrants like OpenAI, Anthropic, and others challenging existing digital giants in the new AI landscape. Moreover, the emerging challenges posed by China in AI further underscore the need to let cloud computing and generative AI markets flourish: as ITIF's China report highlights, while China leads in AI research publications and is catching up with the U.S. in generative AI, its research impact is less significant, and the U.S. should focus on a comprehensive AI strategy that promotes development and adoption rather than containment to sustain its competitive edge.[22]
By deciding not to pursue digital regulation, the ACCC can avoid the problems that are already materializing as a result of the EU's DMA and instead follow the U.S., Taiwan, and Singapore down a path that allows the digital revolution to continue to develop and reap the resultant benefits to both Australia and its consumers. Indeed, amidst a rising China threatening techno-economic dominance, policies that continue to foster innovation in the West and among its core allies have never been more important. Heavy-handed regulation that places significant burdens on America's leading technology firms sends exactly the wrong message about the relationship between the United States and its closest allies in the Indo-Pacific.
Thank you for your consideration.
[1]ACCC, Digital Platform Services Inquiry - March 2025 - Final Report Issues Paper (July 25, 2024), [hereinafter Issues Paper].
[2]The European Parliament and the Council of the European Union adopted the Digital Markets Act on 14 September 2022, a Regulation 2022/1925 on contestable and fair markets in the digital sector and amending Directives (EU) 2019/1937 and (EU) 2020/1828 (DMA) and Regulation (EU) 2022/2065 on a single market for digital services and amending Directive 2000/31/EC (DSA).
[3]See, e.g., Joseph V. Coniglio, Lilla Nóra Kiss, Comments to the Indian Ministry of Corporate Affairs Regarding Digital Competition Law at 9 (May 15, 2024), Comments to the Indian Ministry of Corporate Affairs Regarding Digital Competition Law | ITIF (highlighting how DMA-like legislation has stalled in the U.S. Congress).
[4]See, e.g., Mark Green and Ralph Nader, Economic Regulation vs. Competition, Uncle Same the Monopoly Man, 82 Yale L.J.871 (1973) (discussing how "economic regulation lacks both a comprehensive theory and a consistent goal" and "undermines competition and entrenches monopoly at the public's expense").
[5]Issues Paper at 9.
[7]Issues Paper at 10.
[8]See, e.g., Jean-Charles Rochet & Jean Tirole, Platform Competition in Two-Sided Markets, 1 J. Eur. Econ. Ass'n 990, 991-94 (2003); see also Catherine Tucker, Network Effects and Market Power: What Have We Learned in the Last Decade, 32 Antitrust72, 75-76 (2018); David Evans & Richard Schamlensee, Paying with Plastic: The Digital Revolution in Buying and Borrowing146-47 (2005) ("Multi-homing is common in many multisided industries").
[9]ACCC, Digital Platform Services Inquiry - Report on App Marketplaces at 45 (2021).
[12]Issues Paper at 11-15.
[13]ACCC, Digital Platform Services Inquiry, Report on expanding digital platform ecosystems at 117 (2023).
[14]See, e.g., Eastman Kodak Co. v. Image Technical Services, Inc., et al., 504 U.S. 451, 481, 488 (1992).
[15]See, e.g., Philippe Aghion et al., Competition and Innovation: An Inverted-U Relationship, 120 Q. J. Econ.701 (2005).
[16]Issues Paper at 14.
[17]Daniel Castro, Joseph V. Coniglio, Lilla Nóra Kiss and Aswin Prabhakar: Comments to the European Commission's Directorate General for Competition on Virtual Worlds and Generative AI at 7 (2024), https://itif.org/publications/2024/03/08/comments-to-dg-comp-on-virtual-worlds-and-generative-ai/.
[18]Id.
[19]Issues Paper at 15.
[20]Daniel Castro, Joseph V. Coniglio, Lilla Nóra Kiss and Aswin Prabhakar: Comments to the European Commission's Directorate General for Competition on Virtual Worlds and Generative AI at 8 (2024), https://itif.org/publications/2024/03/08/comments-to-dg-comp-on-virtual-worlds-and-generative-ai/.
[21]See Maureen K. Ohlhausen, Acting Chairman, U.S. Fed. Trade Comm'n, Should We Fear The Things That Go Beep In the Night? Some Initial Thoughts on the Intersection of Antitrust Law and Algorithmic Pricing at 11 (May 23, 2017).
[22]See Hodan Omar, How Innovative Is China in AI?, ITIF (Aug. 26, 2024), How Innovative Is China in AI? | ITIF.