# Connect principles to evidence.

Responsible AI belongs within one versioned governance architecture.

## Published and working methodologies

Published methodology and historical scores remain unchanged. The seven-pillar Responsible AI architecture is a research draft awaiting reconciliation and accountable-owner approval.

## One control, multiple references

Map a control to relevant standards without counting the same evidence twice. A mapping is an interpretation, not certification or a determination of legal compliance.

## An auditable chain

Document the system and scope, applicable requirement, control owner, evidence source, assessment method, reviewer and decision. Preserve revisions and unresolved disputes.

## Explore

- [Responsible AI requires accountable governance.](/responsible-ai)

- [Connect principles to evidence.](/responsible-ai/framework)

- [Choose the assessment that fits the evidence.](/responsible-ai/assessment)

- [Ask for decisions, owners and evidence.](/responsible-ai/board-oversight)

- [Transparency starts with bounded claims.](/responsible-ai/transparency)