Spurred by ChatGPT, South Korea pushes world’s first AI Bill

By Somang Yang

Spurred by the speed and scale of the AI revolution heralded by ChatGPT, South Korea’s National Assembly took a significant step towards passing the world’s first comprehensive bill on Artificial Intelligence (AI).

The “Bill on Fostering AI Industry and Establishing Trust” which passed the Science and Technology Information Broadcasting and Communications Committee, is headed towards a plenary vote by the legislature. The bill mandates the setting of a national strategy (“a Basic Plan”) for advancing the AI industry and fostering AI-specialized companies under the principle of “allow first, regulate later.”

This principle ensures that anyone can research, develop, and launch AI technology and algorithms, but the operator needs to put in place reliability measures if it could have a significant impact on human life, safety, or basic rights. All AI operators must also abide by the AI Ethics Principles.

Under the AI Bill, the Minister of Science and ICT must set out a Basic Plan every 3 years and convene an AI Committee under the Prime Minister’s Office to review policies and investments and allocate budgets. The Committee will be made up of heads of government agencies and experts from the private sector appointed by the President and Prime Minister.

The AI Committee will also host an “AI Trustworthiness Expert Committee” to gather opinions from various fields and conduct discussions and research on fairness, transparency, and other issues.

Additionally, the bill contains measures for the government to designate innovative AI companies to provide systematic support for technology development and commercialization, with the Minister for Science and ICT charged with supporting AI startups, while local and national governments are to support the creation of AI clusters across the country.

Unresolved questions remain

One of the major uncertainties in Korea’s AI policy is how “high-risk” areas will be defined and regulated. Under the bill, high-risk areas are not limited to the obvious, such as autonomous vehicles and nuclear facility management but could cover fields as broad as energy, drinking water, health care, medical devices, transportation, employment and loan review, and also cover data categories such as biometric data. The designation and regulation of AI in high-risk areas fall to the Minister of ICT and the AI Committee, who must set out reliability criteria that operators must meet.

Another unresolved issue is who will lead user protection, given that the Korea Communications Commission (KCC) currently leads on the issue with regard to the internet, while the Personal Information Protection Commission is in charge of protecting privacy and personal information.

As the Ministry of Science and ICT is likely to be in charge of AI ethics enforcement, there is a potential for this issue to develop into an administrative turf war.

When debating the bill, legislators stressed the urgency of creating a legal foundation for AI so that the country would not fall behind its global competitors and asked the administrative branch to coordinate on the unresolved matters outside the legislative process so as to ensure the bill’s swift passage.

Full steam ahead for AI

Meanwhile, outside the passage of the AI Bill, the Yoon administration unveiled further plans to develop the country’s AI capabilities, including support to develop an equivalent to ChatGPT, the AI chatbot acquired by Microsoft. The government will build AI learning data by this year and expand the scope of AI voucher support for the development of innovative AI services, including chatbots. Next year, it plans to promote mainstreaming AI and develop human-centred AI by 2026. The government also said it will be revising copyright laws to enable the use of copyright data to develop regenerative AI, in addition to providing a legal framework for its use. The move came as a part of the “New Growth Strategy 4.0,” an economic development roadmap for 15 future-oriented growth industries ranging from space exploration to self-driving technology, in addition to more support for the semiconductor, rechargeable battery, and display sectors.

When the AI Bill passes the National Assembly Standing Committee, the Legislation and Judiciary Committee, and the plenary session in the first half of 2023, it is expected to result in tangible tax benefits and levy deductions for companies specializing in AI through specific enforcement decrees, and AI-specialized clusters will be designated by region.

Posted in

Related Articles

Indonesia’s Jeopardized Digital Freedoms – A Case Study in Spyware

By Grey Pilarczyk Indonesia has long had a mixed human rights record, particularly regarding the online activity of Indonesians. In recent years, political activists and independent media outlets in the country have faced a barrage of digital threats and cyber-attacks. On May 1st, Amnesty International released a stunning report detailing Indonesia’s use of highly invasive […]

Unpacking the EU AI Act: An ASEAN Perspective

By Nigel Hee The European Union (EU) recently unveiled the Artificial Intelligence Act, a novel piece of legislation that aims to regulate the development, deployment and use of artificial intelligence (AI) systems within the EU. The Act is predicated on a risk-based approach, classifying AI systems into different risk categories and imposing corresponding obligations and […]

Digital Sovereignty in ASEAN

By Mackenzie Gunther With large tech companies owning significant amounts of data, geopolitical tensions, the risk of critical data leaks, and the rising importance of self-reliance in the eyes of world leaders, the concept of who controls data is becoming a high priority.   The global context over the past decade has set the scene for […]