On May 26, 2026, Indonesia's Business Competition Supervisory Commission (KPPU) Chair M. Fanshurullah Asa told the House of Representatives' Commission VI that the country needs a standalone digital markets law. His argument was blunt: digital platforms are no longer mere intermediaries between buyers and sellers but integrated ecosystems folding together logistics, payments, data, algorithms, and AI — and the 1999 competition statute was never built to police them. The KPPU flagged five risks it says existing law cannot reach: abuse of dominance by large platforms, vertical integration and self-preferencing, discriminatory market access, predatory pricing and cross-subsidization, and what it called "anti-competitive practices linked to algorithms and AI."
The regulator's case is not frivolous
It is worth stating the strongest version of the KPPU's argument before disputing any of it. Indonesia's e-commerce market is genuinely concentrated. Shopee alone held roughly 46% of gross merchandise value in 2024, and the top three platforms — Shopee, Tokopedia, and TikTok Shop — controlled about 80% between them, according to Momentum Works data compiled by Databoks. When TikTok's parent ByteDance bought 75.01% of Tokopedia for around $840 million in early 2024, the KPPU's own investigation found a "significant increase in market concentration" and granted only conditional approval in June 2025, with compliance monitoring running until June 17, 2027.
Algorithmic harms are also a real frontier. As far back as 2022, the peer-reviewed Bulletin of Indonesian Economic Studies warned that tacit collusion among sellers using pricing algorithms is a live concern in concentrated, network-effect markets — and that Indonesia's regulators had flagged the difficulty of even detecting it. Asa made the same point in November 2025, arguing that algorithmic collusion can produce uniform pricing "without explicit agreement," which is close to invisible under a statute built around proving a cartel agreement. A self-preferencing ranking tweak or an opaque recommendation engine leaves no smoking-gun memo. That evidentiary gap is the regulator's most defensible motivation.
But the existing toolkit is not empty
Here is where the case for a brand-new statute weakens. The third amendment to Law No. 5 of 1999 is already moving through the legislature and, per analysis from Rajah & Tann Asia, expected to take effect around 2026. That reform is substantial in exactly the areas the KPPU says it lacks. It would shift Indonesia to mandatory pre-merger notification, grant the KPPU search-and-seizure powers to collect digital evidence, formally admit indirect and circumstantial economic evidence, clarify extraterritorial reach over foreign platforms, and — critically — redefine the "relevant market" to recognize that dominance "can arise from misuse of user data, algorithmic discrimination, or AI-based predatory pricing."
In other words, the amendment already targets data-driven and algorithmic dominance. The TikTok–Tokopedia conditions — bans on predatory pricing and self-preferencing, mandated open access to payment and logistics — show the KPPU can extract structural remedies from platforms today, under current law, through case-by-case merger review. The question is not whether Indonesia can act against algorithmic harm. It demonstrably can. The question is whether a second, parallel ex-ante regime adds enough to justify its cost.
The DMA template carries real costs
A standalone digital markets law would almost certainly borrow from the EU's Digital Markets Act: designate large platforms as "gatekeepers" and impose blanket prohibitions — no self-preferencing, mandatory interoperability, data-sharing duties — that apply regardless of whether harm is shown in a specific case. That model is attractive precisely because it sidesteps the evidentiary burden the KPPU complains about. But it does so by presuming guilt from size.
That is a poor fit for Indonesia for three reasons. First, the market is contestable in ways the headline concentration numbers obscure. The KPPU itself acknowledged that the TikTok–Tokopedia merger "does not constitute a significant barrier to entry," and TikTok Shop's surge from a standing start to roughly 11% of GMV in two years is evidence that incumbency is not destiny. Video commerce grew from under 5% of online GMV in 2022 to about 20% in 2025 — disruption, not stasis. Second, many of the practices the DMA bans outright, such as bundling logistics and payments into a platform, deliver genuine efficiencies to the millions of MSMEs that sell through these ecosystems. A blanket prohibition trades those gains for theoretical foreclosure risk. Third, ex-ante rules ossify: they freeze today's business models into law just as the technology shifts beneath them.
Proportionate beats novel
The better path is to let the Law No. 5/1999 amendment do its work and then build narrowly on top of it. If the real gap is detection of algorithmic collusion, the answer is investigative capacity — algorithmic-audit expertise, mandatory logging of pricing-algorithm inputs for firms above a dominance threshold, and the indirect-evidence rules already in the amendment — not a sweeping new prohibition regime. If self-preferencing is the worry, the conduct standard should remain effects-based: demonstrable consumer or rival harm, not mere structural advantage.
The KPPU has correctly diagnosed that algorithms and AI strain a 1999-vintage framework. But Indonesia is mid-stream in modernizing that very framework. Stacking a DMA-style statute on top — before the amended law has even been tested — risks duplicating mandates, multiplying compliance burdens, and signaling to investors that one of Asia's most dynamic digital economies prefers pre-emptive prohibition to evidence-based enforcement. Sharpen the tools already in hand first.