The ValaisOS standard
Governed Intelligence Principles.
Expansive intelligence becomes useful in consequential work when evidence, policy, judgment, and authority remain connected.
What this means
Intelligence can expand without displacing authority.
ValaisOS is being built to expand what a firm can inspect, connect, and prepare, without transferring the firm’s authority to technology. These principles define how evidence, policy, review, and approval remain part of the work.
- 01
The firm remains the authority.
The firm determines what may be accessed, where intelligence may run, who may approve an output, and which artifacts may enter its records workflow.
- 02
Evidence precedes assertion.
Material conclusions should remain connected to sources, versions, assumptions, and relevant context.
- 03
AI output remains proposed until approved.
Intelligence may inspect, interpret, and propose. Accountable authority remains explicit.
- 04
Policy is executable input.
Access, processing, providers, review, retention, and movement into firm systems should follow the firm’s approved operating profile.
- 05
Access is deliberately scoped.
Every identity and connector should receive only the access required for the approved work.
- 06
Approved work enters firm systems deliberately.
An approved artifact should enter an established workflow with its destination, version, approval, and receipt recorded.
- 07
Material activity should be reconstructable.
A firm should be able to understand what information was used, what processing occurred, what changed, who approved it, and where the result went.
- 08
Failure must remain contained.
A stale source, unavailable provider, policy mismatch, or failed connector must not silently convert unreviewed output into authoritative work.
See how these principles shape our regulation-first approach
Advisor Insights applies these ideas to professional practice, including how advisory firms can preserve client context.