Story perspectives

Anthropic Revamps Policy, Prioritizes AI Growth Over Safety

2/25/2026

Summary

Story summary

Anthropic revised its Responsible Scaling Policy, dropping the commitment to halt AI training without guaranteed safety.
The change was prompted by competitive pressures and the evolving AI landscape.
The policy aims to maintain transparency about safety risks while allowing continued development.
Critics, including Chris Painter of METR, warn the shift could raise risks without clear thresholds.
Anthropic co-founder Jared Kaplan labels move pragmatic given market realities and emphasizes transparency.

1 of 1

Story summary

Anthropic revised its Responsible Scaling Policy, dropping the commitment to halt AI training without guaranteed safety.
The change was prompted by competitive pressures and the evolving AI landscape.
The policy aims to maintain transparency about safety risks while allowing continued development.
Critics, including Chris Painter of METR, warn the shift could raise risks without clear thresholds.
Anthropic co-founder Jared Kaplan labels move pragmatic given market realities and emphasizes transparency.