"Concerns Over AI Basic Act's Ambiguous Regulations Hindering Industry Development"

Recently, although the 'AI Basic Act' was enacted to promote the development of the domestic artificial intelligence (AI) industry, concerns have been raised that it may hinder industrial progress contrary to its intended purpose.

On the 31st, the Digital Economy Research Institute of the Korea Internet Corporations Association released an issue paper titled "Promotion Laws That Block Industry: The Paradox of the AI Basic Act." The AI Basic Act, passed by the National Assembly plenary session on the 26th, includes provisions for establishing an AI promotion system, supporting the growth of the AI industry, and creating a foundation based on safety and trust.

The report pointed out that, contrary to the nature of the Basic Act, ambiguous regulations and excessive restrictions raise concerns about hindering the industry.

First, it noted that regulatory ambiguity makes it difficult to determine whether regulations apply. The definition and scope of application of high-impact AI are representative examples. High-impact AI is defined based on the criterion of "potentially having a significant impact on human life, physical safety, and fundamental rights," but there are no specific standards, making it difficult to judge whether it falls under regulation during the early stages of technology development. The report also explained that there are no concrete standards for data quality and fairness criteria, ethical standards, and practical measures.

It also pointed out the possibility of hindering technological innovation. The AI Basic Act stipulates transparency obligations for high-impact AI and generative AI, requiring processes such as risk assessment, certification, and pre-evaluation of impacts. Regarding this, the report analyzed, "This imposes excessive administrative and financial burdens on companies and slows down the speed of technology development and market launch, which, given the fast-changing nature of the industry, can become an obstacle to timely response."

While the government’s regulatory authority is excessive, the report diagnosed a lack of systematic support design. For example, compared to other laws, the requirements for conducting investigations (simple complaints) are excessively lenient, granting the government excessive regulatory power. It pointed out that while regulatory obligations are comprehensive, technical and financial support measures are insufficient.

Finally, the report expressed concerns about weakening global competitiveness. It explained that AI technology standardization criteria are inadequate and that the government’s leading role in establishing these standards could be applied in a way that undermines competitiveness in the global market. Furthermore, while domestic regulations are designed mainly around restrictive frameworks similar to the European Union (EU) AI Act, competing countries such as the United States and Japan adopt relatively autonomous regulatory approaches, which could significantly widen the technological gap between countries.

The report recommended, "A prompt review and amendment of provisions that may hinder industrial development are necessary," and added, "It is essential to clearly assess the reality of the AI hegemonic competition and strategically devise multifaceted measures to avoid falling behind in the competition."