On December 26, Korea passed the Act on the Development of Artificial Intelligence and Establishment of Trust (AI Basic Act). This landmark legislation positions South Korea as the first market in Asia to establish a comprehensive legal framework for AI, following in the footsteps of the EU AI Act.
The AI Basic Act is set to take effect in January 2026, giving businesses one year to prepare.
Tiered approach reminiscent of the EU AI Act
The AI Basic Act follows the suite of core principles that regional observers will have seen crop up in numerous APAC regulatory playbooks and frameworks for AI: ethical AI development, human oversight, responsible use, fundamental rights protection, etc.
Interestingly in terms of narrowing regulation to allow the fostering of innovation, similar to the EU AI Act, the AI Basic Act also adopts a risk-based, tiered approach to regulating AI systems.
Under the AI Basic Act, AI systems are divided into:
- high-impact AI systems which impact “life, bodily safety, and fundamental rights” in activities relating to prescribed sectors (e.g. energy supply, healthcare, medical devices, public services). This parallels with the EU AI Act’s labelling of “high-risk AI systems” and deemed classification within the specified Annex III categories; and
- generative AI systems that mimic input data to generate outputs such as text, sound, images, and other creative content.
At a minimum, businesses adopting high-impact or generative AI systems are required to provide users with advance notice if their products or services incorporate AI elements or are artificially created. Depending on the risk level, particularly for high-impact AI systems, businesses must consider additional compliance obligations including safety and reliability standards, risk management plans, user protection strategies, human oversight, and impact assessments.
Businesses looking to make AI systems available in Korea will need to focus next year on whether each system falls within the categories above and, if so, the most efficient way to meet these requirements before 2026 – whether creating fresh compliance strategies and collateral or adapting those already being developed for the EU or other regulated markets.
Further, non-Korean businesses should note the EU-esque extra-territorial application of the AI Basic Act, requiring them to designate a domestic agent with a presence in Korea where certain thresholds are met.
Non-compliance
Non-compliance with the AI Basic Act may result in fines up to 30 million won (approximately USD23,000 / EUR 20,000), in contrast with the EU AI Act’s penalties of up to EUR 35 million or 7% of worldwide annual turnover for non-compliance. Given the AI Basic Act’s emphasis on nurturing AI technologies, promoting R&D and training, and fostering international cooperation, the relatively lower enforcement regime is unsurprising. However, both laws empower authorities to conduct investigations and impose corrective measures – given the current trend for regulatory intervention in the tech sector in APAC and globally, forward-looking market participants would be well-advised to upskill teams in this area.
On our radar
The parallel establishment of a National AI Committee in Korea marks a significant step forward in the regulation of emerging technologies in that market. This body is expected to roll out further AI policies and initiatives. The Ministry of Science and ICT in South Korea is also committed to accelerating the development of supporting statutes and policies for AI development, all of which will be helpful for businesses to clarify their compliance obligations.
Looking ahead, other jurisdictions like the UK, mainland China, and Japan continue to actively debate AI legislation, with potential new laws projected for 2025.
Our Asia Tech and Data team will be monitoring developments, so watch this space in the new year!
For all those celebrating, all the best for 2025!
*This article is prepared with the aid of reputable translations and secondary sources. We will work with our business partners in Korea to continue to advise our clients on APAC-wide developments in AI. Please see our AI Toolkit for more information.