[The Editors' Verdict] The Dilemma of Fair Competition Regulation and the AI Economy

In September last year, the Korea Fair Trade Commission imposed corrective orders and fines on Naver. It judged that Naver, while operating the ‘Naver Real Estate’ service, unfairly blocked partner real estate agencies from providing property listings to Kakao, thereby unjustly hindering other businesses’ market entry. Naver plans to file an administrative lawsuit, arguing that the measure was to prevent free-riding on the verified property information it independently built at great cost. This case demonstrates various issues that can arise concerning market dominance, the relevance of data, and the use of data as the foundation of artificial intelligence (AI).

Once data is generated, it can be replicated and shared infinitely without cost. Moreover, new data can be created by combining it with other data, which creates incentives to use it exclusively to maximize profits. This makes data openness and sharing difficult, hindering the establishment of an open data ecosystem. Consequently, securing data, an essential input for AI utilization, becomes challenging. It is no coincidence that large internet companies like Facebook and Google, which already have broad customer bases, lead in AI technology utilization. The European Union (EU) is strengthening data sovereignty by recognizing data as a competitive asset and introducing laws and systems to establish a common EU data space and prevent the export of EU data by foreign companies.

Unlike the data ecosystem, the AI ecosystem currently aims for openness. To innovate AI technology development and industrial applications, related companies use collaboration through the openness and sharing of research outcomes as a crucial tool. Google, Facebook, and others release the libraries they develop as open source, allowing programmers to freely utilize them. This openness and collaboration expand the AI ecosystem’s scope, increase the number of companies familiar with leading firms’ technologies, and generate ancillary benefits through the spread of related services. However, this open ecosystem for AI innovation does not remain genuinely fair and open to all participants indefinitely. Although not immediately visible, it is a form with fences beyond the horizon.

For South Korea’s Digital New Deal policy and national AI strategy to succeed amid the clash between the open AI ecosystem and the closed data ecosystem, certain issues must be thoroughly examined. First, a strategy for the complete opening of data to the general public must be established. Beyond policies on personal information protection and utilization, standards are needed to classify and identify data that can be shared and sold without legal restrictions or concerns about trade secret leaks. Sophisticated operational principles and compensation systems for data ownership reflecting industry-specific characteristics must also be established.

Alongside this, standards for data management plans should be developed to improve data quality through systematic enhancements from data generation to collection. To encourage the use of AI open-source algorithms and software, ownership regulations and methods for providing provenance information should be improved. Like data, a compensation system for AI open source must also be in place. Finally, for these systems and frameworks to function properly, an open science culture must be promoted so that citizens can understand and utilize data effectively. A truly open digital ecosystem can only flourish when citizens with data literacy exist, supported by advanced technologies and systems.

Shin Minsu, Professor, Department of Business Administration, Hanyang University

This content was produced with the assistance of AI translation services.