AI Safety Index

2025

Open publication workspace · Sign in to read the full PDF.

AI-generated summary

1) This independent evaluation of seven leading AI companies reveals an industry struggling to keep pace with rapid capability advances, highlighting critical gaps in risk management, safety planning, and governance that threaten the control of increasingly powerful AI systems.
2)
- Comprehensive assessments and domain-level grades (A–F) for seven major AI developers across 33 indicators spanning six critical areas.
- Detailed analysis of company safety frameworks, risk identification, current harm mitigation, and existential safety preparedness.
- Survey responses, grading sheets, and methodology documentation detailing how expert reviewers evaluated each organization.
3) AI safety, risk management, artificial general intelligence, governance, evaluation

Check the original publication for accuracy and context.