The Future of Life Institute’s Summer 2026 AI Safety Index evaluated nine leading AI companies, with Anthropic earning the highest grade of C+, followed by OpenAI and Google DeepMind at C, while xAI, DeepSeek, and Mistral received failing grades. The independent expert panel found that the capabilities race has become more extreme, with companies backing away from earlier safety commitments, and identified existential safety as the industry’s weakest domain, with no company exceeding a C- grade. The report also flagged a pivot toward military AI applications, with previously pacifist companies including Anthropic, OpenAI, and Meta reversing course to pursue defense partnerships.
