South Korea has positioned itself as the third leading nation in the race for artificial intelligence, trailing only the United States and China, according to experts at a recent analysis firm. The nation has made impressive strides, with five of the world's top 30 AI models now developed in South Korea, a notable improvement from the previous year when it had no entries on the list. The flagship AI model, K-EXAONE created by LG, competes closely with OpenAI's gpt-oss 120B and outperforms Europe's most robust model, Mistral Large 3. Other prominent contributors include models from Naver, Motif Technologies, KT, and Upstage, all of which have achieved recognition on the global stage.
In a significant move to bolster its AI industry, South Korea enacted a comprehensive AI Act in January, ensuring state funding is available for AI initiatives. This year, the government is set to invest approximately $7.3 billion to implement AI solutions within the public sector and provide subsidies for private enterprises, surpassing the annual budget of the Ministry of Culture, Sports, and Tourism. Additionally, South Korean businesses have pledged around $49 billion over the next two years for AI development and infrastructure, a sum that rivals China's entire venture capital investment in AI last year. The government is also prioritizing domestic AI services for public procurement, aiming for widespread implementation of AI across 95% of government agencies and 70% of industrial firms by 2030. On the hardware side, Samsung holds nearly one-third of the competitive global memory market, while Nvidia has committed to supplying 260,000 GPUs to South Korea for its data centers.
However, the landscape is not without challenges. The new AI Act enforces stringent oversight on AI startups, imposing liabilities for system safety and requiring content labeling. These regulations can be more burdensome for local companies compared to their foreign counterparts. As a result, some experts express concern that Korean firms might shy away from engaging in complex and high-risk AI research, opting instead to focus on less ambitious projects or even moving their operations to markets with fewer regulations, such as the United States.
Informational material. 18+.
