# Autonomous Regulatory Intelligence (ARI)

## Overview

ARI is the self-governing regulatory AI of DBIS, capable of:
- Detecting system-wide risk
- Modifying AML policies in real time
- Updating FX risk limits
- Triggering liquidity adjustments
- Recommending settlement rule changes

## Architecture

### Cortex Layer

- Regulatory policy generator
- Predictive modeling engine

**Service**: `ari-cortex.service.ts`

**API Endpoints**:
- `POST /api/v1/ari/policy/generate` - Generate regulatory policy
- `POST /api/v1/ari/predict` - Predict risk

### Reflex Layer

- Real-time AML/FX adjustments
- Autonomous sanctions updates
- Automated rule deployment

**Service**: `ari-reflex.service.ts`

### Execution Layer

Interacts with:
- GPN
- M-RTGS
- ALPS liquidity engine
- GASE AML engine

**Service**: `ari-execution.service.ts`

**API Endpoint**: `POST /api/v1/ari/execute`

## Decisioning Example

```
if SARE.FXSP > 0.35:
   tighten_FX_band( SCB )
   reduce_liquidity_limit( SCB )
```

**Service**: `ari-decisioning.service.ts`

**API Endpoint**: `POST /api/v1/ari/decision`

ARI can self-update policies, subject to:
- MSC review windows
- CAA override authority

## Database Models

- `AriPolicy` - Regulatory policies
- `AriDecision` - ARI decisions and actions
- `AriPolicyUpdate` - Policy update history

## Integration Points

- SARE risk engine (Volume V)
- AML velocity engine (existing)
- FX engine (existing)
- GPN, M-RTGS, ALPS, GASE