Bletchley Declaration - Footnote 25

Note: unsupported: The source does not mention seven leading AI companies publishing voluntary safety policies covering nine safety areas including misuse prevention, transparency, and collaboration with government safety institutes. minor_issues: The source mentions that the 16 companies signed 'Frontier AI Safety Commitments' to publish safety frameworks and establish risk thresholds, and commit to refraining from deploying models that exceed their defined safety thresholds. However, it does not explicitly state that they will 'establish risk thresholds'.