Reconstructing Dumping Evidence in Cross-Border E-Commerce Transactions: Law Enforcement Challenges in the Era of Market Algorithms in Indonesia

Photo : DEDI SUPRIADI, S.H.,M.M (Trade Remedy lawyer)

The Indonesian economic landscape has undergone a profound structural transformation characterized by the rapid integration of digital technology into the fabric of trade and commerce. This digitalization has birthed an ecosystem where cross-border e-commerce transactions are increasingly governed by invisible, automated, and highly sophisticated market algorithms. While these technological advancements offer unprecedented efficiency, they have simultaneously introduced complex challenges for law enforcement agencies tasked with maintaining market integrity and protecting domestic industries from unfair trade practices. Specifically, the phenomenon of dumping—traditionally understood as the export of goods at prices lower than their fair value—has evolved into a digital-native form that is difficult to detect, reconstruct, and prosecute under existing legal frameworks. The intersection of international trade law, business competition policy, and computational science now necessitates a fundamental reimagining of how evidence is collected and analyzed in the context of digital-era predatory behavior.

The Dualistic Legal Framework: Dumping versus Predatory Pricing

In the Indonesian legal tradition, the protection of the domestic market from unfair pricing is split between two distinct yet functionally overlapping regimes. The first is international trade law, primarily governed by Law Number 7 of 2014 concerning Trade, which addresses dumping through the Indonesian Anti-Dumping Committee (KADI). Dumping is viewed as a form of price discrimination where a foreign exporter supplies commodities to the Indonesian market at a price significantly lower than the normal value in their home market or a third country. If KADI determines that such practices cause material injury to a domestic industry producing similar goods, the government may impose an Anti-Dumping Import Duty (BMAD) as a corrective measure.

Also read this page: The Global Realignment of Trade Barriers: A Comprehensive Analysis of Tariff Competition and the Sustenance of the WTO Multilateral Framework

The second regime is business competition law, anchored by Law Number 5 of 1999 concerning the Prohibition of Monopolistic Practices and Unfair Business Competition. This framework, enforced by the Business Competition Supervisory Commission (KPPU), addresses “selling at a loss” or predatory pricing. Under Article 20 of Law Number 5 of 1999, business actors are prohibited from supplying goods or services at prices below market value with the intent to eliminate competitors or hinder market entry. While dumping is an international trade violation, predatory pricing is an anti-competitive conduct that falls squarely within the jurisdiction of the KPPU.

Comparative Institutional and Legal Paradigms in Indonesia

Dimension	Anti-Dumping (Trade Law)	Predatory Pricing (Competition Law)
Primary Legislation	Law No. 7 of 2014; Law No. 17 of 2006	Law No. 5 of 1999
Enforcement Body	Indonesian Anti-Dumping Committee (KADI)	Business Competition Supervisory Commission (KPPU)
Core Violation	International price discrimination	Domestic market exclusion via below-cost pricing
Remedial Action	Import duties (BMAD)	Administrative fines, structural remedies
Analytical Approach	Price-to-price comparison (Normal vs. Export)	Rule of Reason; Effect-based analysis
Evidentiary Standard	Material injury to domestic industry	Intent to eliminate; Recoupment probability

The convergence of these two concepts in the e-commerce sector has created a “grey zone” where foreign exporters leverage digital platforms to execute predatory pricing that functions effectively as dumping. This practice initially benefits consumers through rock-bottom prices but poses long-term threats by hollowing out domestic manufacturing and reducing product diversity. The analytical challenge is compounded by the fact that dumping in e-commerce often bypasses traditional physical customs checks through direct-to-consumer models, making it a “ghost” phenomenon that traditional trade monitoring tools are ill-equipped to capture.

The Anatomy of Market Algorithms in the E-Commerce Era

The modern digital marketplace is no longer a static environment where prices are set by human managers in quarterly reviews. Instead, it is a dynamic, high-frequency ecosystem driven by algorithmic pricing—the use of programs to automate the setting of prices based on real-time data analysis. These algorithms are capable of adjusting prices across thousands of SKU (Stock Keeping Units) in milliseconds, responding to competitor price changes, inventory levels, demand surges, and even individual consumer behavior.

Mechanisms of Algorithmic Pricing

Algorithmic systems in cross-border e-commerce utilize several sophisticated machine learning techniques to optimize their pricing strategies. These systems can be differentiated by their degree of autonomy and their functional objectives.

Dynamic Pricing Models: These use real-time market data—including supply and demand trends and competitor prices—to tailor prices dynamically to maximize revenue or market share. For platforms like Amazon or Alibaba, these algorithms calculate the price elasticity of demand for individual consumers, allowing them to adjust prices based on a user’s willingness to pay.
Signaling and Monitoring Algorithms: Some algorithms are designed to track competitors’ online prices and respond instantly. The European Commission has noted that two-thirds of online retailers use such automated software. In a signaling scenario, an algorithm may increase its price to “test” whether a competitor will follow, effectively coordinating market behavior without explicit human communication.
Predictive and Autonomous Learning: Advanced AI models, such as Q-learning algorithms, are capable of discovering optimal pricing strategies through repeated interactions in the market. Computational experiments show that these self-learning algorithms often converge on supra-competitive prices or exclusionary price levels without any intentional human conduct, awareness, or communication between competitors.

Also read this page: Strategic Trade Defense: DSAP Law Firm's Expertise in Antidumping and Safeguard Measures

The proliferation of these tools has created a “Digital Eye” scenario where algorithms operate as autonomous agents, making it increasingly difficult to attribute intent—a critical component of Law Number 5 of 1999—to the human business actors behind the software.

Classification of Pricing Algorithms in Digital Trade

Algorithm Category	Operating Mechanism	Competitive Risk
Heuristic-Based	Uses predefined “if-then” rules (e.g., “be $0.01 cheaper than competitor X”).	Rapid price wars; Easy facilitation of explicit cartels.
Predictive Modeling	Uses neural networks to forecast demand based on historical data and macro factors.	Sophisticated price discrimination; Behavioral manipulation.
Autonomous Learning	Uses reinforcement learning (Q-learning) to maximize long-term profits through trial and error.	Tacit collusion; “Black box” predatory behavior that is hard to explain.
Hub-and-Spoke	Competitors use a common third-party pricing tool that aggregates their data to set prices.	Coordinated price increases across an entire market without direct contact.

Evidentiary Challenges and the “Black Box” Problem

The primary obstacle for Indonesian law enforcement in the era of market algorithms is the “black box” nature of algorithmic decision-making. Traditional competition law and trade law rely on discovering documents, emails, or testimonies that reveal a “meeting of minds” or a clear strategy of selling at a loss.¹⁰ However, when an algorithm “decides” to dump products into the Indonesian market, there may be no paper trail or human instruction to that effect.

The Failure of Traditional Price-Cost Tests

The standard method for proving predatory pricing involves the “Price-Cost Test,” which typically compares the selling price (P) to the Average Variable Cost (AVC) or the Average Total Cost (ATC). In digital trade, this test faces several existential threats:

Unique Cost Structures: Digital platforms often have near-zero marginal costs. Once a software platform or a digital service is built, the cost of serving one additional customer is negligible. This makes the P < AVC threshold effectively useless as a benchmark for predation.
Cross-Subsidization: Global platforms operate in multi-sided markets. They may offer products at a loss in Indonesia to gain data or to lock users into an ecosystem where they make profits through advertising or other services (e.g., fintech).
Algorithmic Opacity: Regulators often lack access to the internal data and logic of the algorithm. Companies frequently cite intellectual property rights to prevent “under the bonnet” technical audits by the KPPU or KADI.

Furthermore, Law Number 5 of 1999 currently lacks clear benchmarks for determining exclusionary intent and recoupment in the digital sector.⁶ Recoupment—the ability of a predator to raise prices later to recover the losses sustained during the predatory period—is theoretically harder to prove in digital markets where barriers to entry are perceived as low, although network effects and “lock-ins” often create high barriers in reality.

Jurisdictional and Technical Barriers

Enforcement is further hindered by the cross-border nature of e-commerce. Foreign digital actors often operate outside the physical jurisdiction of the Republic of Indonesia, complicating the service of legal notices and the collection of evidence. While Article 1 number 5 of Law Number 5 of 1999 defines a business actor as any individual or entity established or domiciled or “carrying out activities” in Indonesia, the practical application of this extraterritoriality has historically been weak.

Barrier Type	Description	Impact on Law Enforcement
Normative	Absence of specific benchmarks for digital “cost” and “intent”.	Legal uncertainty and inconsistent rulings by the KPPU and courts.
Technical	Lack of digital forensic expertise and search/seizure authority.	Inability to secure documentary or digital evidence from foreign platforms.
Jurisdictional	Complexity of prosecuting entities incorporated in foreign jurisdictions.	Enforcement gaps where foreign actors can undercut local MSMEs with impunity.
Transparency	Algorithms are proprietary “black boxes” protected by IP law.	Regulators cannot distinguish between aggressive competition and illegal predation.

Legislative and Regulatory Responses: A Shift Toward Protectionism

Recognizing the existential threat to domestic MSMEs, the Indonesian government has pivot toward a more proactive and protectionist regulatory stance. The most significant development in this regard is the issuance of Minister of Trade (MOT) Regulation Number 31 of 2023, which replaces Regulation Number 50 of 2020. This regulation is specifically designed to curb the influx of cheap imported goods and prevent digital platforms from using their market dominance to stifle local competition.

Key Takeaways from MOT Regulation Number 31 of 2023

The regulation introduces several groundbreaking requirements for cross-border e-commerce operators (PPMSE):

Minimum Price Threshold: Finished goods sold directly from outside Indonesia into the country through e-commerce platforms must have a minimum price of US$100 (FOB) per unit. This is a direct attempt to stop the dumping of low-cost mass-produced consumer goods—such as textiles and accessories—that compete with local artisans.
Separation of Social Media and Commerce: “Social e-commerce” platforms are strictly prohibited from facilitating payment transactions. They are limited to promoting or advertising goods. This was notably the catalyst for the temporary shutdown and eventual merger of TikTok Shop with Tokopedia.
Mandatory Local Presence: Foreign merchants must establish a representative office in Indonesia if they complete more than 1,000 transactions or deliver more than 1,000 packages to Indonesian consumers within a year.
Compliance with National Standards: All merchants, whether onshore or offshore, must display evidence of compliance with Indonesian standards, including SNI (Indonesian National Standard) and Halal certification.

Data Sharing and Surveillance Requirements

MOT Regulation Number 31 of 2023, coupled with customs regulations like Regulation Number 96 of 2023, imposes a mandatory data-sharing obligation on e-commerce platforms. Platforms must share real-time data with the Directorate General of Customs and Excise (DGCE), including:

The seller’s identity and category of goods.
The uniform resource locator (URL) of the goods.
The quantity and transaction currency.
The exact prices and the dates when those prices were in effect.

This data-sharing mechanism is intended to provide KADI and KPPU with the raw data necessary to identify “price manipulation practices” and suspicious dumping patterns that were previously hidden in the private databases of foreign platforms.

Case Analysis: From the TikTok Shop Controversy to the Temu Ban

The practical application of Indonesia’s shifting regulatory landscape is best observed through high-profile cases involving global digital giants. These cases illustrate the government’s commitment to protecting the “64 million MSMEs” that underpin the national economy.

The Ban on Temu (2024)

In October 2024, Indonesia officially blocked the Chinese e-commerce application Temu.13 Temu’s business model—which connects products directly from Chinese factories to Indonesian consumers, effectively skipping all local middlemen—was viewed by the Ministry of Communication and Information and the Ministry of Trade as a threat to domestic commerce.

The rationale for the ban was multi-faceted:

Non-Compliance with MOT 31/2023: Temu’s direct-to-consumer model inherently bypassed the $100 minimum price rule for many of its products.
Predatory Pricing Concerns: The platform’s aggressive marketing and rock-bottom prices were seen as a form of “destructive” pricing intended to wipe out local merchants.
Lack of Legal Entity: Temu was not registered as an electronic system operator (ESO) in Indonesia and had failed to comply with local tax provisions.

The “lifeline” provided by this ban was hailed by Indonesian retailers as a critical step in stemming the tide of “cheap imported goods” that have historically flooded the market.

The TikTok-Tokopedia Merger and Conditional Approval

The saga of TikTok Shop highlights the interplay between regulation and corporate maneuvering. After being forced to stop transactions in September 2023, TikTok acquired a 75.01% stake in Tokopedia for US$840 million to resume its e-commerce operations.

The KPPU’s role in this merger was pivotal. The agency granted conditional approval for the acquisition in early 2024, but with several key requirements designed to mitigate monopoly risks:

Open Access: Maintaining open methods for payment and logistics, ensuring TikTok does not favor its own affiliates.
Prohibition of Predatory Pricing: A strict condition was placed on avoiding predatory pricing practices on the platform.
Monitoring Period: The KPPU will monitor compliance with these conditions until June 17, 2027.

This case demonstrates that the KPPU is increasingly using merger control as a tool for ex-ante regulation in digital markets, setting the stage for longer-term oversight of how algorithms govern platform competition.

Reconstructing Evidence: Forensics and Screening Indicia

Since direct evidence of dumping or predatory intent is often unavailable in algorithmic environments, law enforcement must turn to “indirect” or “circumstantial” evidence. This process involves reconstructing the “logic” of the algorithm through its observable outputs—a technique known as empirical auditing.

Screening Indicia for Collusive and Predatory Behavior

Drawing from international best practices and academic research, several “screening indicia” have been identified that can help the KPPU and KADI detect anti-competitive algorithmic behavior:

Frequency and Speed of Price Changes: High-frequency price adjustments (millisecond intervals) are a primary indicator of automated systems. A seminal study by Chen et al. used the frequency of price changes to identify algorithmic sellers on Amazon.
Price Correlation and “Pegging”: If multiple sellers’ prices are perfectly correlated or “pegged” to the lowest price in the market, it suggests a coordinated algorithmic response.
Structural Break Analysis: Using tests like the “Quandt Likelihood Ratio,” investigators can identify a “structural break”—a sudden change in how a firm reacts to a competitor’s price—which may signal the moment an algorithmic pricing strategy was adopted.
Negative Correlation Between Price and Demand: In healthy markets, prices often rise with demand. However, cartels and predatory actors may keep prices artificially low during high-demand periods to prevent “cheating” by participants or to maximize the exclusionary effect on competitors.
Price Cycling (Edgeworth Cycles): Patterns where prices slowly fall as competitors undercut each other, followed by a sudden jump back to a high level, suggest a cycle of “cheating” and “punishment” in an algorithmic collusion scenario.

Algorithmic Auditing Frameworks

The emerging field of “algorithmic auditing” provides a more structured approach for regulators to “open” the black box. This framework is often categorized based on “Who, What, and When”.

Audit Aspect	Component	Description
Who (Auditee)	Digital Platforms / Large Merchants	Entities meeting specific thresholds for user scale or data control.
Who (Auditor)	First, Second, or Third-Party	Internal assurance, customer audits, or independent regulator-backed audits.
What (Inputs)	Training Data & Features	Assessing whether the data used to train the algorithm includes sensitive competitor info.
What (Outputs)	Statistical Outcome Analysis	Reviewing the actual prices charged to determine if they are discriminatory or exclusionary.
When (Timing)	Ex-ante or Regular Intervals	Audits conducted before a system goes live or as part of periodic regulatory compliance.

The KPPU has proposed that for high-risk algorithms, companies should be required to perform “Mandatory Market Simulations”. This would involve testing the algorithm in a simulated environment that includes a “maverick firm”—an aggressive competitor that does not collude—to see if the algorithm can still converge on competitive outcomes.

The Future of Law Enforcement: 2026 Reforms and Regional Cooperation

The limitations of Law Number 5 of 1999 have prompted a major legislative push for reform. A draft bill to amend the Indonesian Competition Law is currently in the 2026 Priority National Legislative Program. These amendments are designed to modernize Indonesia’s competition framework and provide the KPPU with the institutional muscle needed for the digital age.

Proposed Amendments to the Competition Law

The draft amendment introduces several critical features that address the challenges of market algorithms:

Search and Seizure Authority: The KPPU currently lacks the power to conduct dawn raids. The amendment would grant the authority to conduct searches and seize evidence, enabling the agency to secure digital logs and server data.
Recognition of Indirect Evidence: While currently used in internal proceedings, the amendment would formally recognize economic and communication-based circumstantial evidence as legally valid “clues” (petunjuk) in all court proceedings.
Redefining the “Relevant Market”: The traditional definition of market dominance based on market share is inadequate for digital platforms. The amendment proposes a definition that accounts for data control, network effects, and algorithmic influence.
Strengthened Extraterritoriality: The draft seeks to clarify the KPPU’s authority over foreign entities that have a “direct, substantial, and reasonably foreseeable” effect on the Indonesian market, closing the enforcement gap against offshore dumping.
Pre-Merger Notification: Moving from a post-closing notification regime to a mandatory pre-merger notification system, bringing Indonesia in line with international standards like the EU and US.

ASEAN Digital Economy Framework Agreement (DEFA)

Beyond national borders, Indonesia is a key participant in the ASEAN DEFA negotiations, expected to be concluded by the end of 2026. This framework aims to harmonize digital trade rules across Southeast Asia, including provisions for:

Cross-Border Data Flows: Enabling trusted data sharing while protecting privacy.
Regional Competition Guidelines: Developing the “ASEAN Investigation Manual on Competition Policy and Law (CPL) for the Digital Economy” to provide common standards for AMS (ASEAN Member States) regulators.
Interoperability: Ensuring that diverse digital landscapes across the region do not lead to market fragmentation.

These regional efforts are essential because a single country’s regulation can often be bypassed if the predatory algorithm is hosted in a neighboring jurisdiction with weaker oversight.

Socio-Economic Implications and the MSME Perspective

The debate over dumping and market algorithms in Indonesia is not merely a legal or technical one; it is a battle for the survival of the domestic middle class. MSMEs contribute significantly to Indonesia’s GDP and employment, yet they are the most vulnerable to algorithmic predation.

The Impact on Generation Z and Consumer Behavior

A study on the influence of TikTok Shop’s pricing on Generation Z consumers reveals a “paradoxical effect”. Predatory pricing and “rushed pricing” (flash sales) have a significant positive effect on consumer attraction and impulsive purchasing decisions. However, long-term reliance on such extreme pricing strategies may erode consumer trust and undermine the overall sustainability of the market.

Purchase Urgency: Algorithms create a “fear of missing out” (FOMO) through limited-time offers that drive consumers away from local products that have stable, fair prices.
Erosion of Trust: When consumers realize that the “luxury-style variety for cheap” lacks guaranteed quality, it can lead to a loss of trust in the broader e-commerce ecosystem.

For local producers, the impact is devastating. When foreign exporters use algorithms to price goods below the cost of raw materials in Indonesia, local manufacturers cannot compete even with maximum efficiency. This leads to factory closures and a decline in national manufacturing capabilities.

Summary of MSME Protection Measures in Indonesia

Measure	Legal Basis	Target / Goal
$100 Price Cap	MOT Reg 31/2023	Protecting textile, apparel, and craft MSMEs from ultra-cheap imports.
Social Commerce Ban	MOT Reg 31/2023	Preventing platforms from using social data to dominate the retail market.
Import Duties (BMAD)	KADI / Trade Law	Correcting injuries caused by documented dumping of specific industrial goods.
Task Force Monitoring	Inter-Ministerial Task Force	Cracking down on the illegal entry and circulation of “cheap imported goods”.
Pre-Merger Scrutiny	KPPU / Law 5/1999	Ensuring mergers don’t result in “post-acquisition price increases” or exclusion.

Conclusions: Integrating Law and Technology for Market Integrity

The reconstruction of dumping evidence in cross-border e-commerce represents the new frontier of Indonesian law enforcement. As algorithms become the “invisible hand” of the digital market, the “visible hand” of regulation must adapt with equal sophistication. The current trajectory indicates that Indonesia is moving toward a multidimensional regulatory model that combines strict entry requirements—such as the $100 price cap—with modernized investigative powers for the KPPU.

The success of these efforts will depend on several critical factors:

Forensic Capability: The KPPU and KADI must evolve into “data-first” agencies, capable of performing technical audits and interpreting complex algorithmic outputs as circumstantial evidence.
Legislative Timeliness: The enactment of the 2026 amendments to Law Number 5 of 1999 is essential to provide the legal certainty and institutional authority (search/seizure) needed to prosecute foreign digital actors.
Regional Synchronization: Through ASEAN DEFA and ACAP 2025, Indonesia can ensure that its domestic protection measures are bolstered by regional cooperation, preventing “regulatory arbitrage” where platforms hide behind jurisdictional borders.
Balancing Innovation and Protection: While protecting MSMEs is paramount, regulators must be careful not to create a “digital wall” that stifles the genuine benefits of AI in logistics and personalized commerce.

Ultimately, the era of market algorithms demands a shift from “reactive” law enforcement to “proactive” oversight. By mandating transparency, establishing clear digital benchmarks, and fostering regional collaboration, Indonesia can ensure that its cross-border e-commerce ecosystem remains a space for fair competition rather than a playground for algorithmic predation. The reconstruction of evidence is not just a technical task; it is the fundamental mechanism by which the rule of law is maintained in the digital economy.

Protect your market position from unhealthy algorithmic price wars and dumping; contact us today for specialized strategic advice

Our Contact:

Email : Management@dsaplawfirm.com
Call: 085719442140