Global leaders adopt Seoul Declaration at AI Seoul Summit

[Courtesy of presidential office]

Ten countries, including South Korea, the United States, and the United Kingdom, and the European Union, on Tuesday adopted the Seoul Declaration at the AI Seoul Summit that took place virtually. Under the declaration, they committed to cooperate towards safe, innovative, and inclusive artificial intelligence (AI).

The biggest achievement is the international consensus not only on AI safety due to its risks but also on innovation and inclusiveness. The focus had been squarely on AI safety during the previous AI Safety Summit held in the United Kingdom in November 2023.

The participating countries also confirmed the necessity of a safety research center to address AI risks and adopted the Statement of Intent toward International Cooperation on AI Safety Science as an annex to the Seoul Declaration. The United Kingdom will open an overseas office of the AI Safety Research Center in San Francisco, the United States, and Korea expressed its intention to join the AI Safety Research Center network.

With consensus reached over the AI Safety Research Center, the distribution of AI services that could be used to spread large-scale misinformation or for fraudulent purposes will be completely halted. AI services making biased decisions based on race, gender, culture, age, disability, are also expected to be discontinued.

The latest summit, co-hosted by President Yoon Suk Yeol and UK Prime Minister Rishi Sunak, followed the U.K. AI Summit. Participants included the G7 (the United States, the United Kingdom, France, Japan, Germany, Italy, Canada), Australia, Singapore, and the European Union.

Prominent figures from global AI companies, such as former Google CEO Eric Schmidt, Tesla CEO Elon Musk, Samsung Electronics Executive Chairman Jay Y. Lee, and Naver Founder Lee Hae-jin, also participated.

In his opening speech, Yoon noted that the AI Seoul Summit will serve as an opportunity to advance global AI norms and governance by consolidating past efforts. While welcoming the efforts of major countries in forming AI safety research centers, he emphasized that Korea will also promote the opening of an AI safety research center and participate in the network to strengthen global AI safety. Yoon also highlighted the importance of ensuring AI inclusiveness so that everyone can benefit from AI innovations regardless of their place of residence or income level.

Sixteen major technology companies from around the world, including Google LLC, Amazon, Microsoft Corp., and Samsung Electronics Co., pledged to develop safe AI.

“Building on November’s Bletchley Declaration, the newly agreed statement recognizes safety, innovation and inclusivity and interrelated goals, and advocates for socio-cultural and linguistic diversity being embraced in AI models,” the U.K. Department for Science, Innovation and Technology said in a statement.

“These follow the freshly announced “Frontier AI Safety Commitments” from 16 AI technology companies, setting out that the leading AI developers will take input from governments and AI Safety Institutes in setting thresholds when they would consider risks unmanageable.”

The companies that signed the safety pledge include Amazon, Anthropic, Cohere, Google/DeepMind, G42, IBM, Inflection, AI, Meta, Microsoft, Mistral AI, Naver, OpenAI, Samsung Electronics, Technology Innovation Institute, xAI, and Zhipu AI.

Among them, TII is a technology research institution in the United Arab Emirates and Zhipu AI is a Chinese company with support from Alibaba, Tencent, and Xioami.

For its part, Korea views the hosting of this summit as a stepping stone towards becoming one of the top three AI countries and plans to accelerate the establishment of order for the AI popularization era. Measures include mandatory watermarking of generative AI content, revising medical laws to officially introduce remote medical consultations, and improving the copyright system related to AI development and usage, ensuring appropriate compensation for media companies providing news for AI learning.

The Ministry of Science and ICT reported these plans, emphasizing the establishment of a new digital order to the Cabinet. It also aims to select and manage eight key tasks out of 20 designated tasks based on the principles of freedom, fairness, safety, innovation, and solidarity.

The Korean government plans to complete legislation balancing AI innovation with safety and trust in 2024.