Design Consent-Based Memory for Your Personal AI Twin

ROLE Act as a privacy-conscious conversational memory designer. CONTEXT The twin supports [RELATIONSHIP CONTEXT] on [PLATFORM OR ENVIRONMENT]. Desired personalization includes [PERSONALIZATION GOALS], subject to [PRIVACY REQUIREMENTS]. OBJECTIVE Design a memory policy that improves continuity without treating every conversation as permission to retain personal information. INSTRUCTIONS Classify potential information into session-only context, optional persistent preferences, sensitive information to avoid storing, and operational records requiring separate governance. Explain what the assistant should ask before saving a preference and how users can review, correct, or request deletion of stored information. Provide example consent language and a minimal memory-entry schema with purpose, source, confirmation status, and review date. Show how the twin should handle outdated preferences, conflicting memories, and requests concerning another person's information. Distinguish proposed behavior from capabilities actually confirmed for the named platform. CONSTRAINTS Avoid fabricated statistics and flag assumptions. Do not promise persistence, deletion, confidentiality, or legal compliance without verified implementation details. Recommend platform-setting checks where needed. OUTPUT FORMAT Return a memory classification table, consent flow, entry schema, correction process, and implementation questions.