AI Feature Disclosure
The optional AI layer is disabled until provider, model, retention and contractual decisions are approved.
- Version:
- DRAFT-2026-10-12
- Effective:
- [EFFECTIVE_DATE]
- Updated:
- 2026-10-12
- Language:
- EN
How it works
When enabled, the feature sends a fixed, allowlisted manifest of public market and macro evidence to [AI_MODEL_PROVIDER] through [AI_GATEWAY_PROVIDER]. It does not accept a free-form financial prompt.
Personal accounts, balances, holdings, transactions, trades, tax, housing, identity, email, notes, statements, passwords, secrets and tokens are prohibited evidence fields.
Limits
Output may be incomplete, incorrect or misleading. It must cite supplied evidence identifiers, distinguish observation from interpretation and avoid personalized recommendations or certainty claims.
Provider processing
Retention, training use, abuse monitoring, processing regions and subprocessors are [AI_RETENTION_AND_PROCESSING]. The feature remains disabled until these facts and applicable provider terms are reviewed.
Control
The feature is user-triggered and can be disabled by the operator. Deterministic market observations remain available independently where provider rights allow.