Space Exploration Technologies Corp. Class A Common StockSpaceXAI received an F grade, indicating very poor safety practices and potential regulatory backlash.
Anthropic、Google、OpenAIは、Future of Life Instituteがまとめた最新のAI安全性指数で最高評価を獲得したが、SpaceXAIはF評価を受けた。AnthropicはC+、OpenAIとGoogle DeepMindはそれぞれC、Meta PlatformsはD+、中国企業のZ.aiとAlibaba CloudはともにD-だった。中国のスタートアップDeepSeekとフランスのMistralもF評価だった。この指数は9社のAI企業を6つのカテゴリーで評価し、実存的リスクへの安全性が全体的に最も低いスコアとなった。研究は、業界リーダーがますます自律的になるシステムを制御する信頼できる計画を持っておらず、能力の限界を目指して競争するリーダーたちが以前の安全性へのコミットメントを弱体化または無効にしていると結論づけた。
Space Exploration Technologies Corp. Class A Common StockSpaceXAI received an F grade, indicating very poor safety practices and potential regulatory backlash.
Alphabet Inc Class CGoogle DeepMind received a C grade, indicating weak safety practices, and the report says no leader has credible control plans.
Meta Platforms Inc.Meta Platforms scored a D+, reflecting poor safety practices in the AI Safety Index.
Alibaba Group Holding LtdAlibaba Cloud received a D- grade, indicating weak AI safety practices.
Advanced Micro Devices IncAnthropic received a C+ grade, but the report highlights industry-wide safety failures and lack of control over autonomous systems, which could lead to stricter regulation and reputational harm.
Mistral AI received an F rating, indicating very poor AI safety practices.
OpenAI received a C grade, and the report highlights weakened safety commitments.