Is Korea Ready to Transition into the AI Era Like This? [Reading Science]

-Infrastructure ranks high, policy and ethical preparedness rank low
-Social discussions must begin to establish detailed criteria

Chatbot Iruda. Archive photo.

[Asia Economy Reporter Kim Bong-su] How prepared is South Korea for an artificial intelligence (AI) society? While it ranks among the world’s leaders in hardware aspects such as smartphone and internet penetration, efforts to utilize AI responsibly in terms of policy and ethics remain insufficient. Experts point out that instead of broad declarations, social discussions should begin to establish detailed standards.

◇AI Embedded in Daily Life

In March 2016, Lee Sedol 9-dan, then the world’s top Go player, suffered a crushing defeat to the unknown AI ‘AlphaGo.’ This event became a turning point in spreading awareness about AI. In fact, AI has already deeply penetrated everyday life. Satellite navigation systems (GPS) in cars use AI to find the fastest routes. The soon-to-be-introduced Level 4 autonomous driving technology is impossible without AI. YouTube and Netflix analyze users’ preferences through AI to increase viewing time. AI is even used in pornographic deepfakes. AI is essential in various fields such as email spam filters, language translation, credit card fraud alerts, book and music recommendations, computer virus security, and smart grid technology. Financial transactions, as well as various business, medical, counseling, marketing, and advertising activities, are actively conducted through AI. We are truly in the era of AI becoming a part of daily life. Accordingly, AI ethics are emphasized to prevent misuse and abuse, avoid flawed designs, and prevent personal and social harm caused by unintended negative consequences. The number of papers with titles mentioning AI ethics increased dramatically from single digits annually between 2000 and 2019 to 70.

◇Contemplating ‘Trustworthy AI’

Major countries around the world have responded to these changes. In May 2019, the Organisation for Economic Co-operation and Development (OECD) established the ‘OECD Principles on AI,’ signed by 42 countries. These principles emphasize that AI must be trustworthy and equipped with appropriate safeguards, and that developers and organizations must be held accountable. The European Union (EU) also established ‘Ethics Guidelines for Trustworthy AI’ in April of the same year. In April this year, it proposed regulatory measures based on risk levels through the world’s first ‘Artificial Intelligence Act.’ The United States has been introducing AI regulatory bills since 2019, and big tech companies such as Google and Microsoft (MS) have established voluntary AI development principles. China announced the ‘Guidelines for Building a National Next-Generation AI Standard System’ in August last year. Japan established the ‘Human-Centered AI Social Principles’ in March 2019 and issued guidelines last July.

In South Korea, the controversy over the AI chatbot ‘Iruda’ in December last year underscored the necessity of AI ethics. However, AI ethics at the government level in various countries remain broadly framed without detailed discussions. South Korea, in particular, shows a significant gap between infrastructure and policy/ethics. According to the ‘Government AI Readiness Index’ published by Oxford Insights last year, South Korea ranks 7th among 34 OECD countries in hardware and is first in Asia in terms of data and infrastructure. However, its ranking in measuring how responsibly the government uses AI was only 21st. Jeong Do-beom, principal researcher at the Korea Institute of Science and Technology Information, pointed out, “Governance must be established to comprehensively coordinate AI ethics issues and derive social consensus,” adding, “A practical forum for discussion should be created where all members of society can share, participate, and communicate.”