AI Safety Index Summer 2026: Anthropic leads with C+, OpenAI and DeepMind at C, Meta D+, xAI and Mistral fail

What happened

The Future of Life Institute published its Summer 2026 AI Safety Index — a comprehensive assessment of 9 AI labs across 6 domains: risk assessment, current harms, safety frameworks, existential safety, governance and accountability, and transparency.

Context and impact

Not a single top player earned better than C+ — which FLI interprets as a signal that even the best companies are deeply below the required safety bar given the pace of AI capability advancement. The report arrives as AI systems are being deployed in governments and critical infrastructure.

Details

• Anthropic: C+ — leads in 5 of 6 domains, strong transparency and safety framework
• OpenAI: C — leads in risk assessment (diversity of external testing)
• Google DeepMind: C
• Meta: D+
• Z.ai and Alibaba Cloud: D-
• xAI, DeepSeek, Mistral: F
• FLI conclusion: 'Even top-ranked companies' safety measures are completely inadequate relative to AI capability pace'
• Methodology: 7 independent experts assessed against absolute (not relative) standards