Crypto
AI Giants Falter in Safety Index
The Future of Life Institute's Summer 2026 AI Safety Index has been released, ranking top AI developers on their safety practices. Anthropic topped the list with a C+ grade and a score of 2.66 out of 4.0, while OpenAI trailed slightly with a C grade and a score of 2.28.
The index evaluates companies across six domains and 37 specific indicators, covering transparency, governance, risk assessment, and safety frameworks. Anthropic led in five of the six evaluation domains, with strong marks for transparency, governance, and its Responsible Scaling Policy.
However, a concerning trend emerged: top-ranked firms are quietly walking back previous safety pledges. The report also flagged an increase in military AI collaborations across the industry.