Although criminal penalties are limited to breaches of confidentiality, the main risks in practice lie in three administrative penalty factors: failure to comply with AI disclosure obligations, appointment of domestic representatives, and remedial orders.
series index
Part 1 | Lawyer Hong Seung-kwon “AI has arrived in the legal world”
Part 2 | How should lawyers use AI?
Part 3 | Role division between AI and lawyers
Part 4 | AI legal advice alone is not enough — here’s why lawyers still matter
Part 5 | Lawyer Hong Seung-kwon: South Korea’s AI Law—30 million won fine standard
Hong Seung Kwon
Chief Lawyer, PAN Law Office (skh@law-pan.com)
Consideration of AI framework method
The recently enacted “Framework Act on the Development of Artificial Intelligence and the Establishment of a Foundation of Trust” is significant in that it is the first basic law that recognizes artificial intelligence as an independent regulatory subject in the Korean legal system. This time, I will briefly explain my observations regarding the so-called AI framework law from the perspective of an active lawyer.
This law consists of six chapters. Chapter 1 describes purpose, definitions, basic principles, and scope. As the first comprehensive framework law on artificial intelligence, Article 2, which defines key terms, is particularly important.
The law defines key concepts such as artificial intelligence, AI systems, high-impact AI, generative AI, and AI business operators. High-impact AI refers to AI systems that have the potential to significantly impact human life, physical safety, or fundamental rights. Generative AI refers to AI systems that mimic the structure or characteristics of input data to produce output such as text, audio, images, or video.
In Chapter 2, we will establish a governance system for the healthy development of AI and the construction of a foundation of trust. It provides for the establishment of a National AI Strategy Board under the President and authorizes the establishment of an AI Policy Center and an AI Safety Institute.
Chapter 3 deals with AI technology development and industrial promotion. This includes measures to build and revitalize the AI industry ecosystem, including standardization efforts, training data policies, talent development, startup support, and international cooperation.
Chapter 4 is about AI ethics and trustworthiness. Authorizes the establishment of AI ethical principles by executive order. In addition, if the service utilizes AI, the business operator will be required to disclose the information and will be required to establish a risk management system to ensure the safety of AI.
Chapter 5 includes supplementary provisions such as mandatory fact-finding regarding AI. It provides administrative enforcement tools such as data requests, on-site inspections, and remedial orders, thereby increasing the effectiveness of the law.
Finally, Chapter 6 deals with criminal penalties and administrative fines, perhaps the area of greatest public interest.
Regarding criminal penalties, Article 42 only applies in cases of breach of the confidentiality obligation under Article 7(9) in relation to the National AI Commission. If a committee member, public official, or other official who obtains confidential information in the course of their duties leaks the information or uses it for purposes other than official duties, they may be sentenced to up to three years in prison or a fine of up to 30 million won.
For example, there may be a leak of a company’s AI safety evaluation materials submitted during committee deliberations, or the private misuse of internal information obtained during policy deliberations. It is unlikely that ordinary companies operating AI services will be immediately held criminally liable under this provision.
In contrast, Article 43 provides for administrative fines and is expected to act as a more direct enforcement mechanism in practice. In three situations, an administrative fine of up to 30 million won may be imposed.
First, there is a violation of the obligation of transparency under Article 31(1). If a business provides products or services that utilize highly influential AI or generative AI without notifying users in advance, they may be subject to fines. For example, failing to clearly notify users that they are interacting with an AI chatbot or distributing content generated by generated AI without proper disclosure could result in sanctions.
Secondly, the violation of the obligation to appoint a national representative under Article 36(1). AI businesses that meet certain criteria and do not have an address or office in the country will be required to appoint a local representative. Failure to do so may result in administrative fines.
Thirdly, failure to comply with a suspension order or correction order under Article 40, Paragraph 3. Violations may be subject to fines if the competent authority identifies them through investigation and orders corrective action.
In summary, although the enforcement order detailing the specific provisions of this law has not yet been finalized and clear standards have not been fully established, the enactment of this framework law is of great significance. This sets fundamental regulatory standards in the rapidly evolving AI industry.
We encourage those working in the AI field to reduce uncertainty through proactive inquiries and up-front consultation. We hope this brief overview will help you navigate your new legal landscape.
저작권자 © Korea IT Times
