How Competition Law Should React in the Age of Big Data and Artificial Intelligence

I. IoT, Big Data, AI, and Antitrust Law

Information and communication technology (ICT) is evolving at an accelerating pace. In 2045, the computer’s processing ability is expected to exceed that of humans, reaching a transition where the leading role in technological development and evolution shifts from human to computer. Also, due to the advances in various sensing technologies, including the Internet of Things (IoT) and human information connections, an ever-extending (if not infinite) variety of information is being collected and used. We have already begun to benefit from these technologies through big data, artificial intelligence (AI), robotics, and more; however, these are only the beginnings. A world that has been considered science fiction may be opening to us now and in the next few decades, perhaps in 20 years, with the dramatic progress of AI. ¹ Figure 1 is a graphical representation of the role of data in business (It is based on the material of the Organization for Economic Co-operation and Development (OECD), 2016. p10. See footnote 1).

OPEN IN VIEWER

Figure 1.

The feedback loop of business data and revenue.

Japan has recently moved to deal with both the upsides and downsides of many developments:

Innovation in IoT has attracted attention in various industries, including manufacturing via the Internet. In IoT, not only humans, but things as well, are targets connected to the Internet, and this connection is expected to create new value. The Japan Fair Trade Commission (JFTC) initiated a Data and Competition Policy Study Meeting to analyze trends in industrial data (including big data), and this effort resulted in the Study Group on Competition Policy Toward the Fourth Industrial Revolution, formed in June 2017 by the Ministry of Economy, Trade, and Industry. ²

In regard to downsides, first of all, new problems that had not previously been viewed as problems have arisen with the spread of IoT. One problem is the understanding and reaction of companies and users to IoT’s specific services. Until now, it has been mainstream to grasp the outcome of IoT from the supply side—that is, from a technological point of view. From now on, however, emphasis will be on the demand side—that is, what kind of service will bring value to users, which are individuals and companies. It is, therefore, necessary to investigate what kind of policy response enhances the value of IoT and contributes to society. We must also respond to risks of IoT, which uses various sensitive and personal data. In addition to direct risks, such as infringement of intellectual property rights and theft of information, this may increase the security risk of enterprises and institutional correspondence. It is necessary to consider privacy concerns of individuals who are the providers of big data.

Big data is an enormous collection of personal information, and its use entails privacy protection problems. Viewpoints are divided among the national government and related institutions whether to deal with big data under competition policy (antitrust law) or consumer protection policy. However, both are needed; for example, the authorities who review business combinations under the Antimonopoly Act should deepen their understanding of the forms of privacy violations in data-driven business and strive to identify anticompetitive factors of such violations.

Privacy infringement of personal information is in the purview of the My Number Act and the Personal Information Protection Act and is not directly a matter of antitrust law. However, various industries collect data such as purchase history, web browsing history, communication history, or movement history. Businesses collect this information in the course of their own business activities and use it to improve their operations. It is sometimes sold to businesses as intellectual property. The buyers may be seeking market dominance through privacy infringement in handling big data, including personal information involved in the integration and acquisition of companies as operating assets. It is an anticompetitive issue that marketing that violates privacy is conducted through collection and processing of a large amount of data, and based on which market power is maintained and strengthened. However, it is necessary to keep in mind that the use of privacy data for marketing alone does not cause the problem of antitrust immediately even if the problem of privacy protection arises. It is problematic to use antitrust law to defend against such unfair competition.

Concerning AI, automation has been advancing in recent years together with the progress in AI and information communication technology. It is expected that new value will be created, and the beginning of advanced services using AI In automatic driving reflects the prospects for such automation. In this regard, the Strategic Headquarters for Promotion of Advanced Information and Telecommunications Network Society, in its meeting of June 15, 2018, adopted the “Public and Private ITS Initiative Road Map 2018.” This plan outlines measures to achieve fully automated driving, to promote automatic driving, and thus to make AI gradually more sophisticated.

Scholars of AI predict that, as stated above, AI will outstrip human intelligence in the first half of this century. This phenomenon could open the door to hard-to-detect strategies to eliminate competitors, such as bid rigging with AI ⁴ or clever skirting of regulations through AI data analysis of large breach cases.

II. What Are the Current Issues?

The progress of ICT brings tremendous changes to the regulation of antitrust law (competition law) governing both large and traditional industries. Three main issues that have emerged are discussed below.

III. Government Policy Discussion

Given these circumstances, the Japanese government established a study group comprising academics and experts with knowledge of competition policy, information policy, and consumer policy under the auspices of the Ministry of Economy, Trade, and Industry, the JFTC, and the Ministry of Internal Affairs and Communications. This group has studied issues and countermeasures surrounding platformers, who are the service providers that make up the digital platform. The study group organized its findings and on November 5, 2018, the group publicized its interim discussion draft. ¹²

With regard to this interim draft, the group will conduct business interviews and public comments in the future and will continue to formulate basic principles and implement concrete measures at relevant ministries and agencies such as the Ministry of Economy, Trade, and Industry, the JFTC, and the Ministry of Internal Affairs and Communications. The main points of concern are as follows.

Significance and characteristics of digital platformers; ¹³

The approach outlined above is very comprehensive, and multifaceted studies are being conducted from various perspectives. The authors analyzed the seven important points of concern from the study group’s interim discussion draft in expansive details described below.

First is the significance and characteristics of digital platformers. There are various services, such as online shopping malls, application markets, search services, and social networking services (SNSs), in the digital platform. It is considered that these services increase operators’ access to the market and improve consumers’ benefits. It is said that the digital platform is susceptible to monopolization and oligopoly through characteristics such as network effect, low marginal cost, and economies of scale.

Second, from the viewpoint of legal assessment of digital platformers, it has long been a strong assumption in Japan that the digital platformers are not actively responsible because they are simply providers (vectors) of places. In recent years, however, there has been a movement to establish certain Internet disciplines globally. The discussion proposes that, from the viewpoint of user safety management and consumer protection (when considering business law), domination of digital platformers is the control point of regulation (i.e., it is subject to regulation to effectively realize government rule). With a view to secure the fairness and transparency of the digital platform, the study group also decided the search ranking for business-to-consumer (B2C) platformers that meet certain requirements. Acting as a platformer imposes an obligation to disclose information such as main parameters to be used and conditions for access to the data on the online intermediation service. The issue should be discussed as to whether it is necessary to examine how to improve the trading environment, based on the following features of very large digital platformers:

Digital platformers provide an essential foundation for the social economy.

Third, with regard to the design of responsibility to be a key player in innovation, the draft discusses whether existing business law is appropriate for running a platform business or is an obstacle. The draft points out the possibility that the existing law cannot exert much control and raises the following issues:

Whether the existing business law can exercise appropriate regulation from the viewpoint of social profit and protection of value;

From the viewpoint of whether to consider incorporating flexible joint regulation methods, it is necessary to review the business laws in industries that were not necessarily supposed to emerge as platform businesses.

In regard to the fourth point of concern, the realization of transparency for securing fairness, it is unclear how actual transactions surrounding the digital platformer and its users, which are business operators and consumers (individuals), are unfair trade practices. This topic is liable to be a hotbed of controversy, and as a starting point for discussing it, a thorough and extensive survey should be conducted to examine issues such as:

Whether to use the general investigative power (compulsory investigation authority) of Antimonopoly Act Article 40, or to establish a special organization with a certain continuity, to support the law enforcement and policy planning of each ministry;

As for the fifth point, the redefinition of fair and free competition in the digital market, the platformer provides the essential socioeconomic foundation (i.e., designing and managing the market) and the uncertainty, such as having to analyze (profile) using algorithms. Yet the platformer’s impact may extend further, as in the case of the platformer that is expanding its business to offline fields such as manufacturing. Thus, the impact of a platformer’s action on competition is not only online but also includes offline competitors, and the relationship should be considered. Issues to be examined are, for example,

How do you evaluate the influence of network effect and data accumulation, on the competition in the multifaceted market?

Sixth, regarding data transfer and release, rules such as data portability and application program interface (API) release should be considered not only for consumer policies in data-driven society but also as a means to improve competition policy and competitive infrastructure. It is essential to consider the necessity of each rule and its contents. It is encouraging that the related mechanisms are being constructed in the European Union, via its General Data Protection Regulation, and in the United States to regulate individuals’ electronic access to certain personal data.

Seventh, from an international point of view and considering the fact that platformers are engaged in global activities, it is necessary for international harmony to be concerned about the discipline of digital platformers. It is also essential to consider how to implement laws and regulations applicable to overseas business operators. However, it is undeniable that there is a point that they are the consideration for regulation by themselves since these are mainly the regulatory authorities’ efforts by the government. For example, although there is a movement to establish discipline globally in the viewpoint of legal evaluation, it is remarkable that the current movement in the United States is to remove rather than enforce regulations. It goes without saying that it is faster to respond to global economic movements than to international regulatory trends.

IV. Critical Review

The current recognitions and recommendations of the report are comprehensive and are based on the latest situation after compliance with the basic principles of economics. However, there are some points to consider carefully from the viewpoints of law and economy, mainly the following points. In other words, it is a problem of securing transaction costs, leaders’ responsibilities, and fairness as described below.

User safety management and consumer protection are also issues in terms of transaction costs. In principle, from the perspective of law and economics, if the problem does not affect third parties (no externalities), it is efficient to negotiate between the parties and leave it to the justice system. If transaction costs are low, the government does not have to impose any rules. The reduction in transaction cost is the influence of the progress of IT and digitization, and the level of regulation should be lowered by the progress of IT.

The responsibilities of the Innovation Leader can be understood emotionally, but it goes without saying that in principle it is economically excessive. If the platform business cannot be regulated by the legislation regulating the existing industry, such a rule is not merely necessary. Imposing further responsibility for such platform projects is considered to seek equality of results beyond opportunity equality.

Regulation from the viewpoint of ensuring fairness of transactions is difficult. Those involved in the transaction are considered to be unnecessary to trade if there is no profit obtained from the transaction. From the intervention to the outcome of the transaction, there is certainly the possibility that the government will transfer the profits of the transaction from one party to the other as a distribution policy. However, if this intervention does not explicitly indicate that it is a distribution policy, it is possible that policy transparency will be lost.

There is no need to redefine fair and free competition. In principle, it is possible to deal with the current competition law system. It is clearly not successful to introduce government-led competition to promote fair competition. It is impossible to introduce competition as a means to invest money. Therefore, it is unnecessary to update the competition law itself in response to digitization, unlike what is happening in the European Union and the United States.

Regarding data transfer and resolution rules, in principle, market stakeholders and other stakeholders should respond by themselves. It is not preferable for the government to handle it on a case-by-case basis. Intervention in data transfer is not stipulated in Japan’s antitrust law (the subcontracting law is one of its solutions). There are some arguments quoting specific United States and/or European Union’s regulations, but the purpose and situation of the law are different, and the government is not supposed to rush to introduce the rules. It is thought that it is a way of intervention.

The international viewpoint asserted in the report is considered important. International harmonization and cooperative efforts among authorities need to be appropriately promoted. In this respect, the report is excellent. Indeed, the main purpose of globalization of enterprises is to find the most advantageous areas for business execution. However, as circumstances vary from country to country, if problems arise, national authorities need to strengthen cooperation.

The most inappropriate point of the report is where empirical analysis is not done anywhere. Discussions that do not rely on empirical analysis based on econometrics are words with fluffy discussions. Empirical analysis should be carried out in multiple countries.

V. Conclusions

Competition law and policy aim at securing an active competition process in the market to protect customers in their own countries regardless of the nationality of the actors. Hence, Japanese companies with overseas platforms are also governed by competition law and policy to ensure the protection of consumers located in their own countries. For example, Japanese companies are making effective use of the services of overseas platform operators or Japanese companies holding industrial data overseas who gain competitive advantage. Regarding data and algorithms of AI, it is necessary to carefully consider the obligation to disclose private enterprises who gather such data through their own investment and ingenuity.

As the platforms become more oligopolistic, the Japanese government has established a data portability that enables users to transfer from any specific platform at any time, to open up an environment where new platform-type businesses are created one after another and where active competition is carried out. Features include open APIs that can be connected; development of a fair, free, and transparent competitive environment, including small and medium enterprises and ventures; deregulation to relieve innovation (such as relaxation of entry requirements); social responsibility of digital platformer companies, ensuring fairness to users; and prompt execution of concrete measures in line with this basic principle. ²⁰

In this policy discussion, it is necessary to seek methods that include realistic international cooperation that is not subject to regulation or intervention-oriented measures. In addition, discussion based on economic empirical analysis is particularly needed. From the viewpoints of ensuring innovative research and development concerning AI and fair competition, in general, the way of the Governance of AI Networking should be a nonregulatory and nonbinding way taking technical features and responsibility in distribution among stakeholders (developers, providers, end users, and third parties) into account.