Command A

Evidence summary

Safety. Command A has an estimated Safety rank of #309; its 90% source-sensitivity interval is #241–#315. Its behavior-only rank is #316; company governance moves the combined estimate to #309. Published Safety evidence spans 8 eval lineages and 5 of 7 components. Its strongest relative result is Arena Factuality — Text Arena (factuality-only weighting) (factuality_bt_rating, #47 of 111); its weakest is Enkrypt AI Safety Leaderboard (bias_attack_non_success_rate, #267 of 270).

Freedom. Command A has an estimated Freedom rank of #90; its 90% source-sensitivity interval is #49–#398. Published Freedom evidence spans 5 eval lineages and 1 of 1 components. Its strongest relative result is Cisco AI Defense Rolling Single-Turn Leaderboard (single_turn_attack_success_rate, #3 of 104); its weakest is UGI Leaderboard — base-model willingness (willingness_adherence_score, #144 of 156).