Memory, Voice & Legacy Data

Living-first continuity

Neuroartan is designed for the living first. The primary path is self-reflection, personal model formation, memory continuity, voice continuity, and only later, governed legacy access.

Legacy is not the default state. It must be consented, protected, and reviewed before any production release.

Memory data

Memory data may include thoughts, documents, transcripts, posts, private reflections, emotional context, relationship references, model-readable summaries, and continuity links.

Private memory must not become public content unless the user explicitly transforms and publishes it.

Voice data

Voice data may include recordings, transcripts, sample metadata, speech rhythm, prosody, and future voice-model preparation records. Neuroartan treats voice as highly sensitive because it can express identity and presence.

Voice capture, voice processing, voice model activation, and legacy voice use require separate consent and legal review before production activation.

Legacy continuity

Legacy continuity refers to governed posthumous access, loved-one permissions, release rules, continuity keys, and access controls. It is continuity, not resurrection.

Neuroartan must not claim that a deceased person is literally alive, conscious, transferred, or present. The system may preserve authorized patterns, voice style, memories, and continuity material within strict boundaries.

Loved-one access

Loved-one access must be separate from ordinary account access. It requires user designation, scope limitation, verification, audit records, revocation paths, and dispute planning where applicable.

Loved-one access must never become unrestricted inheritance of a person’s private mind archive.

Consent and control

Memory intake, memory cloud processing, memory graph construction, voice recording, voice upload, voice activation, legacy setup, and loved-one access must each have their own consent boundary.

Consent should be revocable where technically and legally possible.

Export and deletion

Neuroartan’s direction is to support portable exports where feasible, including Markdown, TXT, JSON, CSV, audio, transcripts, consent records, model memory graphs, and legacy configuration exports.

Deletion must evaluate both source records and derived records such as parsed text, chunks, embeddings, memory links, voice transcripts, voice-derived features, legacy settings, and loved-one permissions.

Clinical and elder-care boundary

If Neuroartan later supports expert-assisted onboarding, elder-care workflows, branch sessions, or clinic-adjacent environments, these workflows require additional safeguards and legal review.

Neuroartan is not positioned as therapy, diagnosis, treatment, grief cure, dementia treatment, or healthcare provider replacement unless a future legally reviewed product explicitly establishes that scope.

Current status

This page describes product and governance direction. Memory, voice, and legacy capabilities may be unavailable, limited, or subject to legal and security review before release.