Anthropic leads the Future of Life Institute’s AI safety index with a C+, while OpenAI follows closely with a C. The index, which evaluates major developers across 6 domains and 37 indicators, reveals that no company managed to exceed a C+. Google DeepMind earns a C (2.01 out of 4.0), Meta a D+ (1.67 out of 4.0), while xAI, DeepSeek and Mistral all received F grades. The report finds that safety commitments among leading companies are declining, a concerning trend for the industry. The next index will be released in early 2027.
Source: Read the original article
