South Korea enacts comprehensive law to govern safe use of AI models

South Korea on Thursday formally enacted the world’s first comprehensive law governing the safe use of artificial intelligence (AI) models, according to Yonhap News.

With the Basic Act on the Development of Artificial Intelligence and the Establishment of a Foundation for Trustworthiness now in effect, companies and AI developers are required to take greater responsibility for addressing deepfake content and AI-generated misinformation.

Companies using high-risk AI models, the systems capable of generating content that affects daily life or people’s safety, are required to inform users that their services are AI-based and are responsible for ensuring appropriate safety measures.

Such applications include use in hiring processes, loan assessments and medical advice.

Content generated by AI models is also required to carry watermarks demonstrating its AI-generated nature.

“Applying watermarks to AI-generated content is the minimum safeguard to prevent side effects from the abuse of AI technology, such as deepfake content,” a Science Ministry official said.

Furthermore, the law requires international companies operating in South Korea to designate a local representative if they have global annual revenue of 1 trillion won ($680 million) or more, domestic sales of at least 10 billion won ($6.8 million), or 1 million or more daily users in the country.

OpenAI and Google currently fall under these criteria.

Violations of the act are punishable by fines of up to 30 million won ($20,418). However, the government plans to allow a one-year grace period before imposing penalties to help the private sector adjust.

Under the act, the science minister is required to present a policy blueprint every three years as part of efforts to promote the AI industry.