Scorecard: ailabwatch 2025-09-01 scored Google DeepMind on Boosting safety research = Moderate

NoteThe source text contains a scorecard table with dimensions (rows) and companies (columns). The 'Boosting safety research' row clearly shows: Anthropic 70%, DeepMind 52%, OpenAI 35%, Meta 25%, xAI 0%, Microsoft 13%, DeepSeek 8%. The record claims Google DeepMind scored 52% on 'Boosting safety research' with a letter grade of 'Moderate'. The numeric score of 52% is directly confirmed in the source. The letter grade 'Moderate' is not explicitly stated in the source text provided, but 52% as a raw score is confirmed. The publisher (ailabwatch), entity (DeepMind/Google DeepMind), dimension (Boosting safety research), and scoreRaw (52%) all match the source. The publishedAt date of 2025-09-01 aligns with the source's note 'Up to date as of September 15' and the update note 'as of September 2025', confirming the temporal alignment.