Commitments

This page collects AI companies' commitments relevant to AI safety and extreme risks.

Anthropic
- Responsible Scaling Policy v3.4
- Frontier Safety Roadmap
- Risk reports and system cards
OpenAI
- Preparedness Framework v2
- Frontier Governance Framework
- System cards and safeguards reports
Google / GDM
- Frontier Safety Framework v3.1
- FSF reports and model cards
- Google AI Principles
Microsoft
- Frontier Governance Framework
- Responsible AI Standard
- Transparency reports
Meta
- Advanced AI Scaling Framework v2
- Safety & Preparedness Reports
- White House commitments
xAI
- Frontier Artificial Intelligence Framework
- Model cards
- System prompt repository
Other
- Shared voluntary commitments
- Frontier safety frameworks
- Summit commitments