Home > National > Politics

Google announces establishment of first AI Campus in Korea in talks with President Lee

Published: 27 Apr. 2026, 19:55 Updated: 28 Apr. 2026, 10:21

SARAH KIM
[email protected]

President Lee Jae Myung, right, speaks with Demis Hassabis, CEO of Google DeepMind, on AI cooperation at the Blue House in central Seoul on April 27. [JOINT PRESS CORPS]

Google will establish its first AI Campus in Korea, building a hub for local academia and research institutions to work with global AI experts, the Blue House announced Monday after Demis Hassabis, CEO and co-founder of Google DeepMind, held talks with President Lee Jae Myung.

"Google and Google DeepMind agreed to expand AI cooperation with the Korean research and academic communities," Kim Yong-beom, presidential chief of staff for policy, said in a press briefing at the Blue House in central Seoul.

Kim said it holds "great significance" that Google DeepMind, a British-U.S. AI research laboratory serving as a subsidiary under Alphabet, plans to establish the world's first Google AI Campus in Korea. This excludes Britain, where Google DeepMind's headquarters is located.

Hassabis, the 2024 Nobel laureate in chemistry who often called the "father of AlphaGo," visited Korea for the first time in 10 years after the historic human-versus-computer match of Go, called baduk in Korean, in Seoul to meet with the president, government officials and companies. AlphaGo, an AI system developed by DeepMind, defeated Korean Go champion Lee Se-dol 4–1 in 2016.

The Korean Ministry of Science and ICT earlier on Monday signed a memorandum of understanding with Google DeepMind for an AI-led science and technology innovation project.

The Science Ministry plans to encourage joint research and researcher exchanges centered on the National Science AI Research Center, which begins operations next month, and seek cooperation on development and verification for AI models, the utilization of scientific data and AI bio innovation research hubs. Such collaboration will be carried out with such local institutions as Seoul National University, KAIST and the ministry's three AI bio innovation hubs.

"As our research teams join hands with DeepMind, which possesses world-class scientific AI capabilities, we expect our capabilities in science and technology fields such as bio, weather and climate and future energy to take a significant leap forward," Kim said. "The two sides agreed to cooperate closely on the 'K-Moonshot' project."

The project is a Korean government initiative that integrates AI with science and technology to solve national challenges and accelerate scientific and technological innovation through public-private partnership. By AI resources and research capabilities, it aims to double research productivity by 2030 and tackle 12 national missions across eight strategic fields — including advanced bio, future energy, physical AI, space, materials, semiconductors and quantum technology — by 2035.

Korean Deputy Prime Minister and ICT Minister Bae Kyung-hoon, left, and Google DeepMind CEO Demis Hassabis pose for a photo after signing a memorandum of understanding on AI cooperation at the Four Seasons Hotel in central Seoul April 27. [MINISTRY OF SCIENCE AND ICT]

"Google has decided to open the Google AI Campus in Seoul within this year to significantly expand cooperation with researchers and startups," Kim said, adding that Hassabis has agreed to consider sending a 10-person research team from Google to Korea.

At the beginning of their meeting, President Lee told Hassabis he is "very popular in Korea" due to grandmaster Lee Se-dol's "shocking" defeat in the AlphaGo match 10 years ago.

"He also won a very important game," Hassabis replied. "The only person to defeat AlphaGo."

The president and the DeepMind CEO discussed the responsible use of AI and the need for global guardrails and safeguards. President Lee noted the lack of international control bodies or standards to ensure the safe use of AI.

"My government also has a great interest in AI and is heavily investing in this field, but we are not sure if the development of AI will flow only in the direction of increasing and benefiting the welfare of all people, or if it will develop in the direction of attacking and destroying peace," Lee said.

"It's very important for me to see AI being used for things like science and medicine," Hassabis said, noting he spent his 30-year career "to prove the technology could do something very interesting," like learn how to play Go, but also invent new ideas.

Lee noted that he often uses DeepMind's Gemini, but noted that "sometimes Gemini does work that you have not ordered it to do."

Hassabis said he is "really pleased" that the president can be counted among Google's AI users but noted that such issues could be possible in such foundation models "if your instruction is not specific enough."

"We have many Koreans on the team, so we try to make sure it's very good for the Korean language, for example, so it's great to see the popularity," Hassabis added.

Addressing the challenges with powerful AI systems, the CEO said, "Now we're building agent systems that can do things themselves, and eventually we'll get to AGI [artificial general intelligence]. One of the big risks is to make sure they stick to the guardrails humans give them."

Hassabis noted that establishing international norms is not easy amidst intensifying technological competition between the United States and China, and emphasized the urgent need for cooperation among countries like Korea, Britain and Singapore to establish such safeguards, Kim said in the press briefing.

Lee mentioned the need to prepare for the unemployment and job issues that AI will bring, noting that "now is precisely when basic income is needed," Kim added.

Hassabis agreed on such a necessity, adding that while the state should provide basic services such as housing, education, transportation and tourism, it is also necessary to consider ways to integrate these with capital market principles.

Lee told Hassabis at the end of their talks that "just as the AlphaGo match raised the curtain on the AI era with Korea 10 years ago, I look forward to opening a brilliant future of AI for everyone 10 or 20 years from now together."

Hassabis called Seoul's initiative for international cooperation on AI safety impressive, saying that once a hub is established, it could evolve into a platform to discuss international AI safety standards, Kim said.

A Go board signed by Demis Hassabis, CEO of Google DeepMind, and grandmaster Lee Se-dol, is gifted to President Lee Jae Myung on April 27. [BLUE HOUSE]

Korea is pushing to open a global AI hub in collaboration with international organizations.

Lee asked when AGI might be obtained, and Hassabis said that the technology, capable of utilizing all human cognitive abilities, could become a reality within the next five years, or as early as 2030, according to Kim, which could bring about major social changes exceeding those of the industrial revolution.

Both Lee and Hassabis agreed that while AI can be a powerful tool to solve difficult challenges, it also carries the risk of being used in warfare or exacerbating the wealth gap, calling on the "the international community to pool its wisdom for the responsible use of AI," Kim said.

"Our country possesses semiconductor competitiveness, world-class manufacturing capabilities, stable infrastructure and outstanding talent," Kim said. "These collaborations are proof that Korea's value as a key partner in the AI era is rising, and they are the path to elevating our country's strategic value to an irreplaceable level."

During their meeting, Hassabis gave President Lee a special gift commemorating the 2016 AlphaGo match — a Go board signed by himself and Lee Se-dol.

In October of last year, Lee met with OpenAI CEO Sam Altman, followed by a meeting with Nvidia CEO Jensen Huang in November.

BY SARAH KIM [[email protected]]