Last released Aug 8, 2026
Risk governance for AI outputs and actions: allow / scale / abstain verdicts, CVaR policies over user-defined costs, and auditable, re-verifiable decision records.
Supported by
