OpenAI classified defense agreement

# OpenAI classified defense agreement **Entity class:** Legal, policy, or governance instrument **Collections:** [[collections/War With Empire|War With Empire]] · [[collections/Neurotech|Neurotech]] **Relationship source:** [[research/Interface and Soft Sovereignty Ecosystem Relationship Graph - 2026-09-10|Interface and Soft Sovereignty Ecosystem Relationship Graph — 2026-09-10]] ## Definition **OpenAI classified defense agreement** is represented in the supplied interface and soft-sovereignty relationship graph through the source-qualified relationships below. The entry preserves direction, predicate, evidence key, and epistemic status without treating analogy, adjacency, lexical coincidence, or shared capability as proof of ownership, funding, technology transfer, operational integration, or institutional control.  ## Relationships ### Interface and soft-sovereignty graph patch — 2026-09-10 **Collections:** [[collections/War With Empire|War With Empire]] · [[collections/Neurotech|Neurotech]] **Relationship source:** [[research/Interface and Soft Sovereignty Ecosystem Relationship Graph - 2026-09-10|Interface and Soft Sovereignty Ecosystem Relationship Graph — 2026-09-10]] **Related source articles:** [[articles/The Next Interface Layer|The Next Interface Layer]] · [[articles/The Magic Kingdom and the Managed State|The Magic Kingdom and the Managed State]] #### Outgoing typed edges - **Edge 459 — `red_line` → no mass domestic surveillance** — **VERIFIED**; evidence `OPENAI_DOW`. - **Edge 460 — `red_line` → no direction of autonomous weapons systems** — **VERIFIED**; evidence `OPENAI_DOW`. - **Edge 461 — `red_line` → no high-stakes automated decisions** — **VERIFIED**; evidence `OPENAI_DOW`. - **Edge 462 — `safeguard` → [[wiki/cloud deployment|cloud deployment]]** — **VERIFIED**; evidence `OPENAI_DOW`. - **Edge 463 — `safeguard` → [[wiki/OpenAI-controlled safety stack|OpenAI-controlled safety stack]]** — **VERIFIED**; evidence `OPENAI_DOW`. - **Edge 464 — `safeguard` → cleared OpenAI personnel in the loop** — **VERIFIED**; evidence `OPENAI_DOW`. - **Edge 465 — `safeguard` → [[wiki/contractual protections|contractual protections]]** — **VERIFIED**; evidence `OPENAI_DOW`. - **Edge 466 — `excludes_without_new_agreement` → Department of War intelligence agencies such as NSA** — **VERIFIED**; evidence `OPENAI_DOW`. OpenAI’s March 2 update explicitly states new agreement required. #### Epistemic boundary `VERIFIED`, `VERIFIED_SECONDARY`, `CORPUS_V`, `CORPUS_I`, `CORPUS_H`, `STRONG_TECHNICAL_LINEAGE`, `STRONG_STRUCTURAL_CONVERGENCE`, `ANALYTIC`, and `UNRESOLVED` remain distinct relationship states. `[V]` is documented, `[I]` is structural inference, `[H]` is hypothesis, and `[OQ]` is an open question. No edge establishes ownership, funding, licensing, technology transfer, operational integration, intelligence activity, or institutional control unless its predicate and evidence explicitly do so.   ### Consciousness mapping and transfer graph patch — 2026-09-10 **Collections:** [[collections/Neurotech|Neurotech]] · [[collections/Consciousness Continuity|Consciousness Continuity]] **Relationship source:** [[research/Consciousness Mapping and Transfer Ecosystem Relationship Graph - 2026-09-10|Consciousness Mapping and Transfer Ecosystem Relationship Graph — 2026-09-10]] **Related source articles:** [[articles/Technologies for Consciousness Mapping and Transfer|Technologies for Consciousness Mapping and Transfer]] · [[articles/The Architecture of Continuity and Emerging Neuroinformatics Standards|The Architecture of Continuity and Emerging Neuroinformatics Standards]] · [[articles/2026 Annual Report on Brain-Computer Interfaces|2026 Annual Report on Brain-Computer Interfaces]] · [[articles/The Organic-Synthetic Brain Atlas|The Organic-Synthetic Brain Atlas]] #### Outgoing typed edges - **Edge 528 — `red_line` → no mass domestic surveillance** — **PRIOR_INTERFACE_VERIFIED**; evidence `OPENAI_DOW`. - **Edge 529 — `red_line` → no direction of autonomous weapons systems** — **PRIOR_INTERFACE_VERIFIED**; evidence `OPENAI_DOW`. - **Edge 530 — `red_line` → no high-stakes automated decisions** — **PRIOR_INTERFACE_VERIFIED**; evidence `OPENAI_DOW`. - **Edge 531 — `safeguard` → [[wiki/cloud deployment|cloud deployment]]** — **PRIOR_INTERFACE_VERIFIED**; evidence `OPENAI_DOW`. - **Edge 532 — `safeguard` → [[wiki/OpenAI-controlled safety stack|OpenAI-controlled safety stack]]** — **PRIOR_INTERFACE_VERIFIED**; evidence `OPENAI_DOW`. - **Edge 533 — `safeguard` → cleared OpenAI personnel in the loop** — **PRIOR_INTERFACE_VERIFIED**; evidence `OPENAI_DOW`. - **Edge 534 — `safeguard` → [[wiki/contractual protections|contractual protections]]** — **PRIOR_INTERFACE_VERIFIED**; evidence `OPENAI_DOW`. - **Edge 535 — `excludes_without_new_agreement` → Department of War intelligence agencies such as NSA** — **PRIOR_INTERFACE_VERIFIED**; evidence `OPENAI_DOW`. OpenAI’s March 2 update explicitly states new agreement required. #### Epistemic boundary Primary-source, peer-reviewed, company-claim, preprint, analytic, inherited-corpus, dynamic-metric, source-version-conflict, and unresolved states remain distinct. A current primary source supersedes a stale uncertainty label without erasing the correction history. No edge establishes ownership, funding, licensing, program lineage, technology transfer, operational integration, or institutional control unless its predicate and evidence explicitly do so.