The X-Layer is a suite of six sovereign engines that enforce continuous AI governance between assessments. Every deviation is detected, predicted, explained, unified, sealed, and chained into a mathematically immutable record.
Each engine has a defined standalone role. Together they form a continuous governance intelligence stack.
Captures behavioural baselines from approved system profiles. Compares every live decision against that baseline in real time. When an approved AI system starts behaving differently from what was certified — Sentinel-X knows before any human does.
Reads Sentinel-X data in read-only mode and produces forward-looking failure predictions. The question it answers is not what is happening — that is Sentinel's role. Predictive-X answers what will happen, and how soon.
Takes the combined outputs of Sentinel-X and Predictive-X and translates them into board-level language. Regulators, NEDs, and institutional clients cannot act on deviation scores — they act on narrative. Prism-X produces that narrative.
Combines the outputs of Sentinel-X, Predictive-X, and Prism-X into a single unified intelligence object. Fusion-X produces the answer that no single engine can produce alone — a complete, cross-validated picture of the system's current governance state.
Turns governed decision history into board-level intelligence: situation summary, risk statement, recommendation — written from live Canal data, never from assumption. When the record is insufficient, ANASSA-X says so explicitly rather than narrating confidence that the data does not support.
Chronicling happens within the chain engine: the sealed governance report is generated and permanently appended in one operation. Each report carries a report ID (CHR-XXXXXXXX), a SHA-256 content hash, and a sealed risk verdict with cryptographic chain integrity — the full per-engine record arrives with the Step 3 enrichment. Every governance report produced by Chronicle-Chain-X is appended to a SHA-256 linked chain — one chain per client. Each link contains the hash of the previous link. If anyone tampers with any link in the chain — any record, any field, any timestamp — every subsequent link breaks simultaneously. This is not a promise of integrity. It is a mathematical proof.
Each engine in the X-Layer receives structured input and produces a clean contract for the next. No engine modifies another. The chain of accountability is mathematically preserved.
A VICCA ONE assessment establishes a governance envelope — what the system is certified to do, within what boundaries, under what conditions. That envelope has a validity window. But the system does not stop making decisions while the window counts down.
The X-Layer operates inside that window. It watches every decision. It detects when the system drifts. It predicts when the envelope will be breached. It explains what is happening. And it seals the complete record into a chain that no one can alter.
When the envelope expires or the chain shows critical deviation, VICCA ONE is triggered automatically. The assessment and the enforcement are no longer separate systems. They are the same system.
Every X-Layer engine is governed by an Authority Gate. No engine executes without a valid licence. You activate each layer per client, per system, per tier. Monetisation is built into the architecture.
The X-Layer applies that principle continuously: detected deviations and recorded predictions are written into its hash-chained governance record.