Archetypal / AI Risk Controls
Find the failure. Improve the control.
AI Risk Controls
Use Observe, Score, Enforce, and case review to identify policy gaps, unsupported outputs, and exceptions that need human attention.
01 / AI Risk Controls
Policy Enforcement
Apply approved Guarded rules to allow, warn, rewrite, block, or escalate a supported event, preserving the policy version and reason behind the action.
Explore Policy Enforcement
02 / AI Risk Controls
Precedent
Retrieve prior decisions with their conditions, dissent, and outcomes, then assess whether they apply to the present scenario.
Explore Precedent
03 / AI Risk Controls
Guarded Observe
Connect an inventory of AI applications to authorized activity signals, policy events, and documented monitoring coverage.
Explore Guarded Observe
04 / AI Risk Controls
Assurance Scoring
Compare AI behavior against versioned requirements, inspect failure classes, and retain the context needed to interpret each score.
Explore Assurance Scoring
05 / AI Risk Controls
Activity Attribution
Associate an observed AI event with its application, authorized identity, policy, and source information without overstating what the evidence establishes.
Explore Activity Attribution
06 / AI Risk Controls
Guarded Probe
Connect authorized environment signals to Guarded’s evidence record with explicit permissions, measured coverage, and known observation limits.
Explore Guarded Probe
07 / AI Risk Controls
Decision Evidence
Retain source-linked decision receipts that connect the rule, decisive facts, enforcement outcome, reviewer, and conditions for use.
Explore Decision Evidence
Resources & collaboration
Bring the mission into the design.
Start with the workflow and the decisions it supports. Define what must remain under human authority and what evidence will be needed for review.