Authorities officers and synthetic intelligence trade leaders agreed on Tuesday to use rudimentary safety measures to the fast-growing area and create a world safety analysis community.
Virtually six months after the primary International AI Safety Summit at Bletchley Park in England, the UK and South Korea are internet hosting an AI Safety Summit in Seoul this week. The assembly highlights the brand new challenges and alternatives the world faces with the arrival of synthetic intelligence know-how.
The British authorities introduced on Tuesday a brand new settlement between 10 international locations and the European Union to create a world community just like the UK’s Synthetic Intelligence Safety Institute, the world’s first public group to speed up the event of synthetic intelligence safety science. The community will promote a standard understanding of AI safety and align its work with analysis, requirements and testing. The settlement was signed by Australia, Canada, the EU, France, Germany, Italy, Japan, Singapore, South Korea, Nice Britain and the USA.
On the primary day of the Seoul AI Summit, world leaders and main AI firms gathered for a digital assembly chaired by UK Prime Minister Rishi Sunak and South Korean President Yoon Suk-yeol to debate safety, innovation and AI integration.
In the course of the discussions, the leaders agreed to a broader Seoul Declaration, which emphasizes rising worldwide cooperation in constructing synthetic intelligence to resolve main international challenges, uphold human rights and bridge digital divides world wide, whereas prioritizing being “people-oriented, dependable and accountable. »
“Synthetic intelligence is a massively thrilling know-how and the UK has led the worldwide effort to harness its potential, internet hosting the world’s first AI Safety Summit final yr,” Sunack mentioned in a UK authorities assertion. “However to get the advantages, we now have to ensure it is secure. That is why I am glad that as we speak we bought an settlement to create a community of AI Safety Institutes.”
Simply final month, the UK and the US signed a memorandum of partnership to collaborate on AI security analysis, evaluation and steerage.
The settlement introduced as we speak follows world-first AI safety commitments from 16 AI firms, together with Amazon, Anthropic, Cohere, Google, IBM, Inflection AI, Meta, Microsoft, Mistral AI, Open AI, Samsung Electronics, Know-how Innovation Institute, xAi and Zhipu.ai. (Zhipu.ai is a Chinese language firm backed by Alibaba, Ant and Tencent.)
AI firms, together with these from the US, China and the United Arab Emirates (UAE), have agreed to safety commitments “to not develop or deploy a mannequin or system in any respect until mitigation measures can hold dangers beneath thresholds values”. to a press release from the UK authorities.
“For the primary time on this planet, so many main AI firms from everywhere in the globe have made the identical dedication to AI safety,” Sunak mentioned. “These commitments be sure that the world’s main AI firms are clear and accountable about their plans to develop secure AI.”