2025 AI Safety Index
Summary
An evaluation of seven leading AI companies across six safety and security domains, grading their risk management, governance, and technical safety practices.
Key quotes
The industry is fundamentally unprepared for its own stated goals
none scored above D in Existential Safety planning
self-regulation simply isn’t working, and that the only solution is legally binding safety standards
Claims using this source
The index assesses companies including Anthropic, OpenAI, Google DeepMind, xAI, Meta, Zhipu AI, and DeepSeek. It uses 33 indicators to calculate grades across domains such as Risk Assessment and Existential Safety.