Google, Meta, Microsoft and OpenAI on Tuesday joined more than a dozen other companies at the forefront of developing artificial intelligence to ensure safety at a time when regulators are scrambling to respond to rapid innovation and new risks. He vowed to develop the technology.
The second world conference included companies from China, South Korea and the United Arab Emirates.
These were endorsed by broad declarations from the G7, the EU, Singapore, Australia and South Korea at a virtual meeting hosted by UK Chancellor Rishi Sunak and South Korean President Yoon Seok-youl.
South Korea’s presidential office said the countries agreed to prioritize safety, innovation, and inclusiveness in AI.
“To protect social welfare and democracy, we must ensure the safety of AI,” Yoon said, pointing out concerns about risks such as deepfakes.
Participants noted the importance of interoperability between governance frameworks, plans for a network of safety agencies, and engagement with international organizations to build it. agreement At the first meeting to properly address risks.
Companies also working on safety include China’s Alibaba, Tencent, Meituan, Xiaomi-backed Zhipu.ai, the UAE’s Technology Innovation Institute, Amazon, IBM and Samsung Electronics.
They have committed to publishing a safety framework for measuring risk, avoiding models that do not adequately mitigate risk, and ensuring governance and transparency.
In response to the declaration, Beth Burns, founder of METR, an organization that promotes the safety of AI models, said, “We need international agreement on a ‘red line’ where AI development becomes unacceptably dangerous to public safety. That is extremely important.”
Computer scientist Yoshua Bengio, known as the “godfather of AI,” welcomed the efforts, but noted that voluntary efforts must be accompanied by regulation.
Since November, the debate over AI regulation has shifted from long-term doomsday scenarios to more pragmatic concerns, such as how AI will be used in areas such as healthcare and finance, a leading language model said. Aidan Gomez, co-founder of the company Kohia, said on the sidelines of the conference. Summit.
China, which co-signed the Bletchley Agreement to collectively manage AI risks at its first meeting in November, did not attend Tuesday’s meeting but is expected to attend an in-person ministerial meeting on Wednesday. A South Korean presidential official said.
AI industry leaders such as Tesla’s Elon Musk, former Google CEO Eric Schmidt, and Samsung Electronics Chairman Jay Y. Lee attended the conference.
The next meeting is scheduled to be held in France, officials said.
