Back
FLI AI Safety Index Summer 2025
webCredibility Rating
3/5
Good(3)Good quality. Reputable source with community review or editorial standards, but less rigorous than peer-reviewed venues.
Rating inherited from publication venue: Future of Life Institute
Published by the Future of Life Institute, this index provides a structured external audit of major AI labs' safety practices, useful for tracking industry accountability trends and identifying gaps between stated safety commitments and measurable actions.
Metadata
Importance: 72/100organizational reportanalysis
Summary
The Future of Life Institute's AI Safety Index Summer 2025 systematically evaluates leading AI companies on safety practices, finding widespread deficiencies across risk management, transparency, and existential safety planning. Anthropic receives the highest grade of C+, indicating that even the best-performing company falls significantly short of adequate safety standards. The report serves as a comparative benchmark for industry accountability.
Key Points
- •Anthropic leads all evaluated companies with a C+ grade, while most other major AI labs score significantly lower across critical safety domains.
- •Widespread inadequacies found in risk management frameworks, existential safety planning, and safety governance structures across the industry.
- •The index evaluates companies across multiple safety dimensions, providing a structured comparative assessment of industry-wide safety performance.
- •Results suggest the AI industry as a whole is failing to meet adequate safety standards even as capabilities rapidly advance.
- •The report functions as an accountability tool, applying external pressure on labs to improve safety practices through public comparative scoring.
Review
The Future of Life Institute's AI Safety Index provides a comprehensive evaluation of seven leading AI companies' safety practices, revealing critical systemic weaknesses in responsible AI development. The assessment spans six domains: Risk Assessment, Current Harms, Safety Frameworks, Existential Safety, Governance & Accountability, and Information Sharing, with independent expert reviewers conducting rigorous evaluations. The report's most alarming finding is the fundamental disconnect between companies' ambitious AI development goals and their minimal safety preparations. Despite claims of approaching artificial general intelligence (AGI) within the decade, no company scored above a D in Existential Safety planning. This suggests a profound lack of coherent risk management strategies, with companies racing toward potentially transformative technologies without adequate safeguards. The index highlights the urgent need for external regulation, independent oversight, and a more systematic approach to identifying and mitigating potential catastrophic risks.
Cited by 26 pages
Cached Content Preview
HTTP 200Fetched Apr 9, 202679 KB
2025 AI Safety Index - Future of Life Institute
Skip to content AI Safety Index - Winter 2025
Latest edition Featured in: Semafor, Reuters, Axios, Quartz, NBC News, Mashable, Euro News, LA Times, and more. December 2025 Future of Life Institute AI Safety Index
Summer 2025
AI experts rate leading AI companies on key safety and security domains. 17 July 2025 Full report PDF 2-Page Summary Coverage Scorecard Company Company grade & score Anthropic C+ 2.64 OpenAI C 2.10 Google Deepmind C- 1.76 x.AI D 1.23 Meta D 1.06 Zhipu AI F 0.62 DeepSeek F 0.37 Domains Breakdown Hint: View this webpage on desktop for a visual overview of scores across all domains Risk Assessment 5 indicators ↗
Current Harms 5 indicators ↗
Safety Frameworks 5 indicators ↗
Existential Safety 5 indicators ↗
Governance & Accountability 5 indicators ↗
Information Sharing 5 indicators ↗
Survey Responses OpenAI ZhipuAI x.AI Overall Grade Overall Score Anthropic C+ 2.64 OpenAI C 2.10 Google Deepmind C- 1.76 x.AI D 1.23 Meta D 1.06 Zhipu AI F 0.62 DeepSeek F 0.37 Domains Hint : Click on a domain to inspect Risk Assessment 6 indicators ↗
Current Harms 8 indicators ↗ Safety Frameworks 4 indicators ↗ Existential Safety 4 indicators ↗ Governance & Accountability 5 indicators ↗ Information Sharing 6 indicators ↗ Survey Responses ↗ C+ B- C D A- A- - C B C F C- B C- C+ D+ D- D B - F D+ D+ F C- C+ D D+ D+ F D- D - F D F F D+ D F D- F F D+ F - How are these scores calculated? See the methodology. Max Tegmark on the AI Safety Index
FLI's President Max Tegmark discusses the importance of driving a 'race to the top' for safety amongst AI companies. Video • 03:30 Key findings
Top takeaways from the index findings: Anthropic gets the best overall grade (C+) The firm led on risk assessments, conducting the only human participant bio-risk trials, excelled in privacy by not training on user data, conducted world-leading alignment research, delivered strong safety benchmark performance, and demonstrated governance commitment through its Public Benefit Corporation structure and proactive risk communication.
OpenAI secured second place ahead of Google DeepMind OpenAI distinguished itself as the only company to publish its whistleblowing policy, outlined a more robust risk management approach in its safety framework, and assessed risks on pre-mitigation models. The company also shared more details on external model evaluations, provided a detailed model specification, regularly disclosed instances of malicious misuse, and engaged comprehensively with the AI Safety Index survey.
The industry is fundamentally unprepared for its own stated goals Companies claim they will achieve artificial general intelligence (AGI) within the decade, yet none scored above D in Existential Safety planning. One reviewer called this disconnect "deeply disturbing," noting that despite racing toward human-level AI, "none of the companies has anything like a coherent, actionable plan" for ensuring such systems re
... (truncated, 79 KB total)Resource ID:
df46edd6fa2078d1 | Stable ID: sid_k5c9RuNZ8K