The South Korean government has proceeded to enforce the "AI Basic Act," aiming to establish itself as a top 3 global AI power. The law's key feature mandates human oversight in high-risk AI sectors and requires clear labeling of AI-generated content. Meanwhile, startup companies complain that vague regulations and compliance burdens may hinder their competitive edge.
South Korea began enforcing the "AI Basic Act" today (22 Jan 2024), describing it as the world's first comprehensive AI regulatory framework. The law aims to enhance trust and safety in the AI sector and position the country as a leader in this technology. However, startup groups have expressed concerns that compliance burdens may obstruct their growth.
This law takes effect ahead of similar efforts in Europe, where the EU AI Act will be phased in gradually through 2027 amid global debates over AI regulation. The United States favors a light-touch approach to avoid stifling innovation, while China has started issuing some rules and proposed establishing a global AI regulatory coordination mechanism.
The core of South Korea's law requires companies to implement human oversight over AI systems with "high impact," covering areas such as nuclear safety, drinking water production, transportation, healthcare, and finance, including credit assessment and loan screening.
Additionally, companies must notify users in advance if AI is used in services and clearly label AI-generated outputs, especially when they are indistinguishable from real content. Penalties are severe; for example, failure to label AI content can result in fines up to 30 million won (approximately 641,000 baht).
The Ministry of Science and ICT stated that this legal framework is designed to promote AI adoption alongside establishing safety and trust foundations. The draft law underwent extensive consultations, and businesses will receive at least a one-year grace period before administrative penalties are enforced.
Baek Kyung-hoon, Minister of Science and ICT and former head of AI research at LG, said the law represents an "important institutional foundation" toward South Korea's goal of becoming one of the world's top three AI powers.
However, the Startup Alliance expressed strong concerns and dissatisfaction over unclear key details. They noted that some legal provisions remain ambiguous, possibly leading companies to adopt overly cautious approaches to avoid regulatory risks, which could stifle innovation. They also questioned, "Why must South Korea rush to be the first country to do this?" amid intense global competition.
The ministry indicated plans to establish advisory platforms and support centers during the grace period, consider measures to reduce industry burdens, and possibly extend the grace period if domestic and international conditions allow.
/sourceReuters