Global tech giants pledge safety commitments at AI summit in Seoul

Korea and Britain secured commitments from 16 global AI tech companies to a set of safety outcomes at a global AI safety summit in Seoul on Tuesday.

President Yoon Suk Yeol speaks in a virtual leaders’ session of the AI Seoul Summit at the Blue House in central Seoul Tuesday. [PRESIDENTIAL OFFICE]

Korea and Britain secured commitments from 16 global AI tech companies to a set of safety outcomes at a global AI safety summit in Seoul on Tuesday.

Google, Meta, Microsoft, Amazon, OpenAI, IBM, Anthropic and Samsung Electronics were among the leading tech companies that agreed to new "Frontier AI Safety Commitments" at the AI Seoul Summit, according to the British government in a statement.

The new commitments to step up AI safely were agreed on by AI tech companies from countries including the United States and China, the world's two biggest AI powers.

China's Zhipu.ai was among the companies that signed onto the commitments, along with the United Arab Emirates' Technology Innovation Institute, France's Mistral AI and Korea's Naver.

The two-day summit brings together global leaders, tech industry representatives and experts in virtual and in-person sessions to address risks posed by AI and present a framework for global AI governance. It followed up on the inaugural AI Safety Summit held in Bletchley Park in England last November.

The 16 tech companies will each publish safety frameworks on how they will measure the risks of their frontier AI models and examine the risk of misuse of technology by bad actors, according to the British government.

The frameworks will also outline when severe risks would be "deemed intolerable" and what these companies will do to ensure such thresholds are not surpassed.

The companies further committed to "not develop or deploy a model or system at all" if mitigation measures can't keep risks below the thresholds in more extreme circumstances.

President Yoon Suk Yeol and British Prime Minister Rishi Sunak co-hosted the virtual leaders' session Tuesday evening, as Seoul aimed to bring innovation and inclusivity to build on the dialogue started last year.

“The AI Seoul Summit will serve as an opportunity to consolidate our efforts and advance AI norms and governance on a global scale,” Yoon said in his opening remarks from the Blue House in central Seoul. “First, we must ensure the safety of AI to minimize its potential negative impacts and protect the well-being and democracy of our society.”

He highlighted that Korea plans to establish an AI Safety Institute to join a global network and encouraged countries to pursue innovation through free and open research and development. AI innovations, he said, can serve as “new growth drivers for the global economy” and help to address shared challenges like environmental pollution.

Yoon said that it is “important to ensure the inclusiveness of AI so that everyone can enjoy the benefits of AI regardless of where they live and how much they earn” and said he hopes the summit “will serve as a stepping stone for global solidarity.”

The session was attended by the Group of Seven (G7) countries — the United States, Canada, France, Germany, Italy and Japan — along with Singapore and Australia. These countries also attended last year's AI summit.

A screen shows an announcement of the AI Seoul Summit in downtown Seoul Tuesday. [AP/YONHAP]

"It's a world first to have so many leading AI companies from so many different parts of the globe all agreeing to the same commitments on AI safety," Sunak said in a press statement Tuesday. "These commitments ensure the world's leading AI companies will provide transparency and accountability on their plans to develop safe AI."

He said it further "sets a precedent for global standards on AI safety that will unlock the benefits of this transformative technology."

Building on the Bletchley summit, Sunak highlighted that "together with the Republic of Korea, we are continuing that success by delivering concrete progress at the AI Seoul Summit."

The leaders signed the Seoul Declaration for safe, innovative and inclusive AI at the summit, underscoring their shared dedication to building on international cooperation and dialogue on AI "in the face of its unprecedented advancements and the impact on our economies and societies."

They recognized that "Al safety, innovation and inclusivity are inter-related goals" that are important to encompass in international discussions on Al governance to "address the broad spectrum of opportunities and challenges that the design, development, deployment, and use of Al presents and may present."

The leaders agreed to support those creating or expanding on Al safety institutes, research programs and other relevant institutions, including supervisory bodies.

They also issued a Seoul statement of intent toward international cooperation on AI safety science. They agreed to advance a common international scientific understanding of AI safety and support state-backed institutions.

A ministerial meeting and an AI Global Forum involving key industry and academia leaders will be held on Wednesday.

"Ensuring AI safety is crucial for sustaining recent remarkable advancements in AI technology, including generative AI, and for maximizing AI opportunities and benefits, but this cannot be achieved by the efforts of a single country or company alone," Korean Minister of Science and ICT Lee Jong-ho said in a statement.

He stressed expectations for companies to "implement effective safety measures throughout the entire AI lifecycle of design, development, deployment and use."

France will host the AI Action Summit in early 2025.

BY SARAH KIM [kim.sarah@joongang.co.kr]