Anthropic Removes Explicit AI Development Pause Commitment from Updated Responsible Scaling Policy

Summary

Anthropic has published Version 3.0 of its Responsible Scaling Policy (RSP), removing the explicit commitment to pause training of more powerful AI models if safety measures fail to keep pace with capability advances. In previous RSP versions, Anthropic's AI Safety Levels (ASL) framework included language stating the company would 'temporarily pause training of more powerful models if our AI scaling outstrips our ability to comply with the necessary safety procedures.' The updated policy replace