In a significant move towards ensuring the safe development and deployment of artificial intelligence, leading tech firms including OpenAI, Amazon, Microsoft, and Google DeepMind have pledged to establish comprehensive AI safety frameworks.
These measures aim to mitigate or entirely prevent potential harms associated with AI technology.
This voluntary commitment, also supported by Chinese firm Zhipu.ai and the UAE’s Technology Innovation Institute, was announced at the AI Seoul Summit.
This event serves as a follow-up to the Bletchley AI Safety Summit, co-hosted by the UK and the Republic of Korea.
It emphasizes the importance of international cooperation in establishing global AI safety standards.
The companies have collectively agreed that if risks associated with AI cannot be sufficiently mitigated under extreme circumstances, they will refrain from developing or deploying such models or systems. This proactive stance highlights their commitment to prioritizing safety over advancement in high-risk scenarios.
Detailed Safety Frameworks
Participating companies have committed to publishing detailed safety frameworks. These documents will outline the methodologies for assessing risks associated with their AI models, including identifying and preventing severe risks deemed intolerable. This transparency is expected to set a benchmark for AI safety protocols worldwide.
“These commitments ensure the world’s leading AI companies will provide transparency and accountability on their plans to develop safe AI,” stated UK Prime Minister Rishi Sunak. He emphasized that this agreement sets a global standard for AI safety, essential for unlocking the transformative benefits of this technology.
Continuing the Momentum from Bletchley Park
This recent agreement builds on the momentum from the Bletchley Declaration, where 27 nations committed to collaborating on AI safety measures. It also aligns with last November’s summit at Bletchley Park, where “like-minded countries and AI companies” agreed to conduct safety tests on AI models before their release. Notably, Google DeepMind stands out as the only major AI lab that has allowed the UK’s AI Safety Institute to perform pre-deployment safety tests.
The commitment from these tech giants to establish AI safety frameworks is a pivotal step towards managing the risks associated with artificial intelligence. This collaborative effort underscores the global consensus on the need for robust AI safety measures to ensure the technology’s benefits are fully realized without compromising safety.
+34 91 991 95 65
+61 410 497229