Abstract
As intelligent information technologies including AI have changed our society as a whole, a paradigm shift toward an intelligent information society has gradually accelerated. Such technologies have been used in a diverse way, depending on target areas and purposes, making the nature and degree of resulting risks inconsistent. Data-based AI systems that are widely utilized these days use a vast amount of personal data from information subjects as their main materials, thereby raising concerns about privacy infringements. At the same time, the issue of bias coming from the process of data input and output reveals new types of infringements on legal interests. Moreover, in both public and private spheres, decision-making processes automated via algorithms replace and complement the thought and behavior of mankind, creating a third realm that is not explained by the existing dual structure characterized by rights-duties, actions-responsibilities, and causes-effects.In connection with the benefits and risks of the Intelligent Revolution led by the 4th Industrial Revolution, sufficient data to create a social consensus have yet to be accumulated identified. Therefore, it is not easy to establish norms for predicting and preventing risks on the premise of a clear order of value. This means that setting up specific legal means to protect intelligent information technologies is hardly expected to effectively control risks in this rapidly changing intelligent information environment. Against this back , this study is designed to come up with ways to improve legal systems, which meet requirements for the social acceptance of such technologies including economy, laws, politics, and public spheres. In particular, this paper focuses on establishing legal systems that will successfully deal with risks and conflicts caused by intelligent information technologies, among the aforementioned requirements.
At the 20th and 21st National Assemblies, a number of revised bills were proposed, whose main aim is to establish intelligent information-related regulations, support, and governance. The Framework Act on Intelligent Informatization currently in operation contains a wide variety of regulations on social impact assessment, safety measures, the establishment of dedicated organizations, and so forth but lacks specific enforcement ordinances guidelines for effective implementation. Above all, it does not include special regulations on discrimination, unfairness, and other right infringements stemming from biased AI algorithms. As a result, current legal systems are hardly forecast to effectively respond to a diversity of risks brought about by an intelligent information society. Over the recent several years, active discussions on legal systems have been conducted to nimbly react to the intelligent information environment based on AI. In relation to leading legal policies, this study reviews the cases of the EU, Germany, the UK, and the USA. These nations have somewhat differently balanced innovation promotion and risk control in accordance with the level and use of technologies but all of them are found to have willingness to set up risk management-based regulatory systems to respond to potential risks.
Like the requirements for social acceptance, national strategies in an intelligent information society should be devised in a multidimensional way. The key strategy for the acceptance of an intelligent information society is to formulate legal systems containing innovative norms such as risk management, AI ethics, supportive regulation, risk measurement/assessment, and so on. The principles of self-control, behavior-centered control, and proactive/current control should be first established in order to be equipped with risk control systems that do not block the acceptance of intelligent information technologies. Legal systems should also be converted, focusing on risk management, principle-based regulation, and co-regulation, in order to effectively control the uncertainties of risks. Moreover, intelligent regulation systems for the Intelligent Revolution should be effectively built by revising legal systems via accreditation, impact assessment, autonomous validation, and transparency measures and by setting up governance based on consumer protection and stability management.