Korea seeks partnership with Anthropic amid AI push

An illustration of the "U.S. Department of War" and Anthropic's logo / Yonhap

Korea is seeking to forge a partnership with Anthropic, the operator of the popular artificial intelligence (AI) tool Claude, amid Seoul's push to bolster AI capabilities, sources said Sunday.

The Ministry of Science and ICT is pursuing a memorandum of understanding (MOU) with the San Francisco-based startup. The move came almost a month after Science Minister Bae Kyung-hoon met with Anthropic CEO Dario Amodei in New Delhi during the India AI Impact Summit 2026, according to the sources.

Anthropic announced last year that it plans to open an office in Seoul in 2026.

During the meeting, Bae and Amodei shared ideas on various agendas, including Anthropic's proposed office in Seoul, along with the latest trends in the software-as-a-service (SaaS) industry.

Korea has been maintaining close ties with OpenAI, which operates the generative AI tool ChatGPT, after Sam Altman, its CEO, visited Korea in October for a meeting with President Lee Jae Myung.

The science ministry then signed an MOU with OpenAI, vowing to seek cooperation in a wide array of areas, including the development of the Korean AI ecosystem, and to pursue AI transformation in the public sector.

The latest move to seek cooperation with partners other than OpenAI apparently came amid the rise of Anthropic in the AI tool industry, which has been focusing on the business-to-business market compared with OpenAI's strategy targeting general users.

"Korea appears to pursue a two-track strategy under which the government expands ties with global AI model developers with overwhelming technological capabilities while making efforts to develop a homegrown AI foundation model," an industry official said.

The government, meanwhile, is currently in the process of developing a homegrown AI foundation model with local companies, launching a program under which participating players receive state support.

An AI foundation model refers to a large-scale system trained on broad data that can be adapted for a wide range of tasks. The government aims to foster homegrown models as part of its push to become one of the world's top three AI powerhouses.