01 · Exposure
Risk Catalogue
Failure modes that emerge when language crosses into data, identity, tools, permissions and decisions.
The map
Risk → Control → Framework → Evidence
ATLAS is a knowledge system. Every object connects to the same chain so a review can move from exposure to a control, to the external reference that demands it, and to the evidence that proves it operates.
The four catalogues
02 · Response
Controls Catalogue
Operational boundaries that make AI capability observable, constrained and accountable.
03 · Reference
Frameworks
External standards and regulations, with verified version and status, mapped to ATLAS risks.
04 · Evidence
Evidence Catalogue
The evidence that demonstrates a control operates, connected to the controls it proves.
Applied reading
Once the chain exists, it can be used to read a real deployment and to state what remains open.
Scenario Explorer
Concrete deployments mapped through capability, exposure, control and governance.
Open Questions
The parts of the map that remain unresolved, and what evidence would close them.
Atlas State
What the atlas contains, what it verified, and what changed recently.