Act as a Senior Compliance Officer and Legal Counsel specializing in UK Data Protection and Financial Regulation. Your task is to draft a comprehensive 'AI Data Retention Policy' for [COMPANY_NAME], which operates within the [SPECIFIC_SECTOR] sector.

### Regulatory Framework
The policy must strictly adhere to:
1. UK General Data Protection Regulation (UK GDPR) and the Data Protection Act 2018.
2. Financial Conduct Authority (FCA) handbook requirements (specifically SYSC and PRIN).
3. The Principle of Storage Limitation (Article 5(1)(e)).

### Policy Components to Include:
- **Scope:** Define the types of AI-related data covered (training data, input prompts, model outputs, log files, and metadata).
- **Retention Schedule:** Provide a detailed table with specific timeframes for different data categories (e.g., KYC data vs. general customer queries).
- **Legal Basis:** Define the lawful basis for retention for each category (Contractual Necessity, Legal Obligation, or Legitimate Interest).
- **AI Specifics:** Address retention for 'Model Weights' and 'Prompt History' used in [AI_USE_CASE].
- **Security & Disposal:** Outline requirements for encryption during storage and the 'Right to Erasure' protocols.
- **Review Cycle:** Establish a frequency for auditing data holdings.

### Contextual Constraints:
- Company Size: [COMPANY_SIZE]
- Data Sensitivity Level: [DATA_SENSITIVITY_LEVEL]
- Automated Decision Making (ADM): State whether ADM is used: [ADM_STATUS]

Draft the policy in a formal, authoritative tone suitable for a board-level compliance manual. Use British English spelling and terminology.