Government officials and AI industry executives agreed Tuesday to implement basic security measures in the rapidly changing field and establish an international security research network.
Nearly six months after the first global summit on AI safety in Bletchley Park in England, Britain and South Korea will host the AI safety summit in Seoul this week. The meeting underlines the new challenges and opportunities the world faces with the advent of AI technology.
The British government announced this on Tuesday a new agreement among ten countries and the European Union to establish an international network similar to the UK's AI Safety Institute, the world's first government-backed organization, to accelerate the advancement of AI safety science. The network will promote a common understanding of AI safety and align its work across research, standards and testing. Australia, Canada, the EU, France, Germany, Italy, Japan, Singapore, South Korea, the UK and the US have signed the agreement.
On the first day of the AI Summit in Seoul, world leaders and leading AI companies gathered for a virtual meeting chaired by British Prime Minister Rishi Sunak and South Korean President Yoon Suk Yeol to discuss AI security, innovation and inclusion .
During the discussions, leaders agreed on the broader point Seoul Declarationemphasizing increased international cooperation in building AI to tackle major global problems, uphold human rights, and bridge digital gaps worldwide, all while prioritizing “human-centered, trustworthy, and responsible.”
“AI is a hugely exciting technology – and Britain has led global efforts to address its potential, hosting the world's first AI Safety Summit last year,” Sunak said in a UK government statement. “But to get the benefit, we have to make sure it's safe. That is why I am pleased that today we have reached an agreement for a network of AI safety institutes.”
Just last month, Britain and the US signed a partnership memorandum of understanding to collaborate on research, safety evaluation and guidance on AI safety.
The agreement announced today follows The world's first AI safety commitments from 16 companies involved in AI, including Amazon, Anthropic, Cohere, Google, IBM, Inflection AI, Meta, Microsoft, Mistral AI, Open AI, Samsung Electronics, Technology Innovation Institute, xAi and Zhipu.ai. (Zhipu.ai is a Chinese company backed by Alibaba, Ant and Tencent.)
The AI companies, including those from the US, China and the United Arab Emirates (UAE), have agreed to safety obligations “not to develop or deploy any model or system at all if mitigations cannot keep risks below thresholds” , according to the British government statement.
“It is a world first that so many leading AI companies from so many different parts of the world all agree on the same AI safety obligations,” Sunak said. “These commitments ensure that the world's leading AI companies will provide transparency and accountability about their plans to develop safe AI.”
