AI Output Review Policy

1. Principle

The first draft, table, code, image, or automation flow created with AI is a starting point. Customers and experts must review facts, rights, security, privacy, and business fitness before final use.

2. Sensitive Data

Resident registration numbers, passwords, API keys, raw customer personal data, and company confidential information must not be included in requests, messages, files, or deliverables.

3. Customer Review

Before accepting a delivery, the customer must confirm that it meets the requested scope and review criteria. Acceptance may trigger a mock payout ledger or future real settlement flow.

4. Expert Responsibility

Experts must not pass through AI-assisted output without review. They should verify, edit, check sources, and confirm practical execution quality.

5. Future Notices

Additional rules will be added when external LLM providers, code execution, deployment automation, or copyright review tools are connected.