Claude agents have achieved a significant milestone by operating autonomously for over 45 minutes at peak efficiency, nearly doubling their performance since October 2025. This improvement is accompanied by a surge in user trust, with experienced developers now auto-approving 40% of actions—up from 20%. These advancements align with Anthropic’s focus on AI safety and controllability, reinforcing the reliability of Claude agents in managing complex tasks independently, including navigating desktops and executing multi-step processes.
Claude agents operate autonomously for 45 minutes, boosting user trust
