Second Thoughts: Key Jurisdictions Smarten Up on DMA

For more than a decade, experts have described a “Brussels Effect,” in which jurisdictions worldwide follow the European Union’s (EU) lead on government intervention. In recent years in tech markets, that effect pits the EU’s “precautionary principle”—regulating against risks before they materialize—against the cost-benefit approach other jurisdictions have traditionally followed. With the Digital Markets Act (DMA), the Brussels Effect at first looked like an unchallengeable juggernaut: jurisdictions across several continents almost immediately began drafting their own versions. But as the DMA’s results roll in, would-be adopters are looking past the Brussels Effect to the law’s true costs, and the DMA-curious—from Japan to Singapore—are producing refreshingly measured policy.

The European Commission (EC or Commission) gave itself a glowing review at the DMA’s two-year mark, declaring implementation a success without grappling with the law’s broader failure. Independent analyses have been less charitable. Last year, ACT gathered seven of our EU- and United Kingdom (UK)-based members to discuss the DMA with American legislators, who were visiting Brussels on an official Congressional delegation trip. Our members’ complaints ranged from mandates forcing their preferred distribution platforms to adopt trust- and security-killing vulnerabilities to compliance-driven delays in access to newer artificial intelligence (AI) features. On that last point, ACT surveyed small tech firms more broadly and estimated the delays cost an individual company $109,000 to $528,000 per year—significant sums for startups and micro-enterprises.

Other critiques measure the DMA’s record and its inherent ex ante structure against its animating purposes: fairness, contestability, and consumer choice. Innovators Network Fellow Satya Marar argues that the DMA’s broad prohibitions and mandates undermine those objectives. On contestability, for example, he notes that a dominant platform is “less vulnerable to challenge if mandates and restrictions reduce the degree to which platforms can differentiate user experiences” along dimensions such as security or privacy. Dirk Auer of the International Center for Law and Economics likewise refutes the EC’s report card, showing that its cited successes are mostly compliance measures that fail to benefit consumers or leave them worse off.

That evidence has an audience. Asia-Pacific (APAC) countries are laser-focused on digital competition, and many at first seemed poised to enact DMA-style frameworks without a second thought. That risk remains—and we recently evaluated the level for key jurisdictions in the region—but the DMA’s downsides are sinking in, and several of these same countries are applying the brakes on, or altogether abandoning, their DMA efforts.

Singapore. Singapore has examined platform competition closely and has repeatedly concluded that it does not need an ex ante regime. The Competition and Consumer Commission completed its E-commerce Platforms Market Study in 2020, finding that the agency’s existing ex post competition framework is more than sufficient. The study recommended interpretive guidance rather than new rules, which the agency later issued in 2022 to address market definition for multi-sided platforms and the assessment of market power in the digital era. Since then, Singaporean policymakers have turned digital platform ex ante regulation proposals away, and Singapore continues to regulate platforms for online harms and infrastructure resilience while leaving market power to its general competition law.

Taiwan. Taiwan has rejected the DMA approach explicitly as well. The Taiwan Fair Trade Commission (TFTC) enforces the Taiwan Fair Trade Act, a flexible framework prohibiting unfair trade and competition practices, similar to the United States’ own Federal Trade Commission (FTC) Act. Applying it to digital markets, the TFTC published a 2022 white paper outlining its enforcement principles. As Acting TFTC Chairperson Andy Chen put it in a 2025 interview, the paper concluded that most digital competition issues could be addressed “within the existing analytical framework,” given sharpened digital enforcement skills. No DMA-inspired framework is pending before the Taiwanese legislature.

Japan. Japan’s Mobile Software Competition Act (MSCA) is clearly based on the DMA, but its scope is narrower—smartphones and app distribution—and, most importantly, its allowances for privacy and security are far more robust. The Japan Fair Trade Commission (JFTC) actively engaged stakeholders throughout implementation to understand how the EC balanced security and privacy against openness. It concluded that the EC never reconciled those goals and parted with the DMA approach to prioritize privacy and security as a benefit for smartphone users.

Australia. The Australian Treasury proposed ex ante digital competition rules resembling the DMA in December 2024, and the Australian Competition and Consumer Commission (ACCC) closed its comment period on developing proposed legislation on February 14, 2025. Draft legislative text was expected in late 2025 or early 2026, but has yet to appear. From late 2024 to late 2025, Australia looked to be moving inexorably toward a copied-and-pasted DMA framework, but we are hopeful that interest in rushing ahead with a DMA copy is waning. In an interview at the Digital Futures Summit in Australia, ACCC Commissioner Luke Woodward confirmed that Australia would not be taking as broad a brush as the DMA as it considers a framework for digital markets.

Republic of Korea. Even at the Korean Fair Trade Commission (KFTC), sentiments seem to be shifting. In recent remarks on a panel in Singapore, the KFTC’s special investigations director signaled a preference for ex post enforcement of competition law in digital markets. He noted that “[t]he trade-off, of course, is that case-by-case enforcement requires proving market dominance and competitive harms . . . We think that cost is worth paying to avoid overregulating the sector where the competitive dynamics are still changing.” The latest thinking, therefore, appears to be a contrast from the previous momentum behind novel frameworks based on the DMA—including the Online Platform Fairness Act, which is not a pure copy, but would impose broad prohibitions and mandates similar to DMA’s.

Thailand. A constellation of Thai agencies—chiefly the Trade Competition Commission of Thailand (TCCT)—have been studying digital markets to evaluate potential policy changes. TCCT issued its market report on e-marketplace businesses in September 2025, followed by its Guidelines on Multi-Sided Platforms and E-Commerce Businesses, and set up a digital platform subcommittee to monitor and prevent unfair trade practices on digital marketplaces. In parallel, the executive branch has weighed a far more sweeping measure, the Digital Platform Economy Act (DPEA), closely mirroring the DMA and the EU’s Digital Services Act (DSA). But Thailand’s Minister for Digital Economy and Society recently signaled that the draft will be revised to reflect Thailand’s different circumstances and the DMA’s lack of success, noting that “[w]e originally based it on EU law, but I instructed officials to slow down and revise it because it was too strict. We need to find the right balance for Thailand.” That is thoughtful skepticism about whether the DMA model fits the country.

The DMA’s spread has not stopped, and several APAC jurisdictions could still reverse course. But the trend matters to the small and medium-sized enterprises—including the growing domestic base of entrepreneurs—that absorb these frameworks’ costs. Every jurisdiction that studies the DMA’s record before copying it is one where small innovators and startups keep access to the tools, platforms, and security guarantees their businesses depend on. That is not a rejection of competition enforcement in digital markets—it is a sign that the Brussels Effect is waning and thoughtful regulators are discovering that DMA-style regulation is not the competition panacea the Commission claims it is.