Responsible and Ethical AI Use

QUILL includes optional AI-assisted features. These features are designed to be user-controlled and review-first.

Responsible-use policy

1. AI output is advisory, not authoritative.
2. You must review all generated output before applying, publishing, or sharing it.
3. You are fully responsible for downstream use of AI-generated content.
4. Do not use AI features to violate privacy, law, policy, or intellectual-property rights.
5. Avoid sending sensitive, regulated, or confidential data to cloud providers unless you are explicitly authorized.

Trust controls in QUILL

• explicit user-initiated AI actions
• approval flow before document-changing actions
• plain-language connection and status messaging
• startup trust and consent acknowledgement
• secure key handling via Windows Credential Manager (with encrypted fallback)

Provider boundaries

When a cloud provider is selected, prompts you send are processed under that provider's terms and retention policy. QUILL cannot override provider-side policy behavior.

For highest privacy, prefer local model backends when available.