Eight pillars. One capability spine.

What the pillars are

The 8 Pillars name the eight institutional capabilities a legal function must develop to operate AI at scale. Unlike a maturity model (which describes how far a function has moved on each capability) or the 6-Layer Architecture (which describes the cadence at which each capability operates), the pillars describe what the function must be capable of at all. The Module Library is organised against the pillars: every one of the 58 modules has a primary-pillar anchor, and the Free Baseline Diagnostic scores the function against all eight. The pillars are load-bearing canon — they have been stable across multiple methodology cycles and are unlikely to require structural revision.

How the pillars work

Capabilities, not categories.

The pillars are not a taxonomy of AI use cases. They are the institutional capabilities the function must develop in order to deploy any AI capability at all. Strategy comes first not because it is more important than the others, but because without it the other seven drift toward whatever the loudest stakeholder or the most accessible vendor pulls them toward. Governance is fourth not because it is fourth in priority, but because it sits between the capabilities the function builds internally (P1–P3) and the capabilities the function deploys externally (P5–P8).

Each pillar has its own detail page that names the canonical sub-capabilities, the modules that operationalise them, and the diagnostic dimensions that score the function's posture. The diagnostic outputs a band placement per pillar; the advisory engagement reads against the band map; the Module Library is the inventory of available work to close the gaps.