Longterm Wiki

Responsible Scaling Policies

Responsible Scaling Policies (RSPs) are voluntary commitments by AI labs to pause scaling when capability or safety thresholds are crossed. As of December 2025, 20 companies have published policies, though SaferAI grades the three major frameworks 1.9-2.2/5 for specificity.

Related

Related Wiki Pages

Top Related Pages

Safety Research

Anthropic Core Views

Risks

Cyberweapons RiskBioweapons Risk

Analysis

AI Safety Intervention Effectiveness MatrixIntervention Timing Windows

Approaches

Dangerous Capability Evaluations

Key Debates

AI Safety Solution CruxesCorporate Influence on AI PolicyWhy Alignment Might Be Hard

Concepts

Model RegistriesSelf-Improvement and Recursive EnhancementLong-Horizon Autonomous Tasks

Policy

AI Whistleblower ProtectionsVoluntary AI Safety Commitments

Other

Red TeamingDario AmodeiElon MuskAI Evaluations

Historical

International AI Safety Summit Series

Sources

Tags

responsible-scalingvoluntary-commitmentssafety-thresholdsfrontier-labsthird-party-evaluation