South Korea's AI Legislation

On December 26, 2024, South Korea made a significant leap in artificial intelligence (AI) governance by passing the "Basic Act on the Development of Artificial Intelligence and the Establishment of Trust" (AI Basic Act) during the National Assembly's plenary session, culminating a four-year process of discussions and legislative refinement. This makes South Korea one of the early adopters of comprehensive AI legislation, following the European Union's AI Act.

Key Highlights of the AI Basic Act

To provide a concise summary, the following highlights focus on the Act's most impactful provisions without repeating detailed elaborations covered later.

• Ethical AI Principles: Transparency, safety, fairness, and respect for human dignity, supported by an 'AI Ethics Charter.'

• High-Impact AI Regulation: Oversight of critical AI systems in healthcare, law enforcement, and public safety.

• National AI Committee: A body to monitor compliance and draft strategies every three years.

• Transparency Requirements: Disclosure of decision-making logic, data sources, and AI-generated content.

• Support for Innovation: Funding for startups, collaborations, and international partnerships.

• Risk Management Protocols: Proactive risk assessments for high-stakes applications.

Key Provisions of the AI Basic Act

1. National AI Basic Plan and Governance: The Act allows the Minister of Science and ICT to establish a National AI Basic Plan every three years to enhance national AI competitiveness (Article 6). It also provides a legal basis for the National AI Committee, established to oversee compliance and policy implementation (Article 7), and the AI Safety Research Institute, focused on protecting citizens from AI-related risks (Article 12).

2. Promoting AI Industry and Innovation: To foster a thriving AI ecosystem, the Act outlines government support for R&D (Article 13), standardization (Article 14), policies for learning data (Article 15), and AI utilization (Article 16). It further emphasizes AI clusters (Article 23), AI data center policies (Article 25), and special support for startups and SMEs (Articles 17 and 18).

3. Ensuring Safety and Preventing Risks: High-impact AI and generative AI are identified as regulatory priorities. Operators are required to ensure transparency (Article 31), safety (Article 32), and compliance with obligations for voluntary safety verification (Article 30). This includes private sector-led assessments to evaluate AI systems' reliability and safety before deployment, providing a framework to address potential risks proactively. The Act also mandates AI impact assessments (Article 35).

Elaborating on Key Concepts

1. Ethical AI Principles and the AI Ethics Charter: The "AI Ethics Charter" is a cornerstone of the legislation, aimed at fostering trust in AI technologies. It emphasizes transparency and fairness while safeguarding human dignity, ensuring that AI systems align with societal values. By codifying these principles, South Korea seeks to set a global benchmark for ethical AI deployment.

2. High-Impact AI Regulation: By categorizing certain systems as "High-Impact AI," the legislation ensures stringent oversight of technologies that significantly affect human life. These systems must meet rigorous safety and transparency requirements, especially in sectors like healthcare and law enforcement. For example, AI used in medical diagnostics will undergo thorough evaluations to prevent potential harm.

3. National AI Committee’s Role: The committee’s responsibilities extend beyond strategy formulation. It serves as a watchdog, ensuring compliance with safety and ethical standards, and acts as an advisory body to guide AI-related policymaking.

4. Transparency and Accountability: Transparency requirements compel providers to disclose how AI systems operate, including details such as decision-making logic and data sources, empowering users to make informed decisions. For instance, generative AI must indicate that its outputs are machine-generated, empowering users with clear information.

5. Encouraging Innovation: Through Supportive Policies To foster innovation, the legislation introduces funding opportunities, encourages cross-sector collaborations, and positions South Korea as a global leader in AI research and development. Regional innovation hubs and dedicated grants are among the key initiatives.

6. Safeguarding Against Risks: Mandatory risk management protocols ensure that businesses proactively address potential risks, especially in high-stakes applications such as autonomous vehicles, where failures could result in significant harm, or financial systems, where errors might lead to widespread economic disruptions. These protocols are vital to maintaining public trust in AI technologies, such as those used in autonomous vehicles or financial systems.

Broader Implications

Minister Yoo Sang-im highlighted the significance of this milestone, stating, "The passage of the AI Basic Act establishes a foundation for national AI competitiveness. It marks a critical step for South Korea to position itself as a global AI leader, enabling large-scale investments and alleviating uncertainties for corporations." The Act is expected to boost public and private investments while enhancing global partnerships.

Implementation Timeline

The AI Basic Act will take effect in January 2026, following a one-year preparatory period for subordinate regulations and guidelines, which are expected to include standards for AI safety verification, detailed protocols for transparency, and guidelines for fostering public-private partnerships in AI development. This ensures that the law integrates smoothly into the market and is fully operational.

With the passage of this landmark legislation, South Korea has firmly positioned itself at the forefront of AI governance, setting an example for other nations to follow.