Anthropic Calls for Global Pause on Advanced AI Development

Odaily Planet Daily report: Anthropic has called on leading AI labs to consider slowing the pace of development, noting that AI systems are evolving at an extremely rapid rate and may soon be capable of self-improvement without human intervention—a capability that could pose significant societal risks. Internal data disclosed by Anthropic shows that the performance of its most advanced models is rapidly improving, with AI progress appearing to move toward “recursive self-improvement,” in which AI systems autonomously optimize themselves without human input. Anthropic states that it would be globally beneficial if the world could choose to slow down or temporarily pause frontier AI research, allowing social structures and related studies to catch up with technological advancements. Anthropic proposes establishing a global agreement that clearly defines how to slow AI development and implements mechanisms to verify whether competitors are complying with the agreement.