NITT v1.0 – Truth-in-Labeling Standard for Digital Identity (github.com)

NITT v1.0, published by Spark on 2025-10-28 (canonical ID NITT-DIG-GOV-STD-v1.0), is a public “Truth‑in‑Labeling” standard for any claimed mind upload, teleport, exocortex or digital-immortality protocol. It mandates an explicit disclosure sentence — “This procedure will terminate the original process and create one or more new persons (branches) who are psychologically continuous with the original at capture. CI < 1 for any realized protocol. Claims of ‘teleport,’ ‘survival,’ or ε = 0 are prohibited.” — and requires audited metrics and consent procedures before any system may use survival‑framing language. The repo records a cross‑model debate outcome in which two AIs conceded that non‑branching survival is physically unattainable, so commercial or research claims of “you survived” would be deceptive. For practitioners and policymakers the standard matters because it reframes legal, ethical and technical responsibilities: branches are new persons (no automatic inheritance of debts, votes, licenses), mandatory audit rights allow demand for CI/energy/timing proofs, and a Public Energy Ledger ties high‑fidelity simulation to real thermodynamic costs. NITT specifies measurable compliance targets (example thresholds: epsilon_delay_ms < 10 ms, epsilon_state < 0.02, epsilon_loop_h_inf ≤ 1.5) and insists CI < 1 even when those are met (i.e., “high‑fidelity branch,” not survival). The standard thus forces clearer consent language, verifiable continuity metrics, and honest marketing — shifting how AI/ML labs, product teams, and regulators must disclose and audit identity‑transfer claims.