extract: 2026-03-08-intercept-openai-trust-us-surveillance

Pentagon-Agent: Epimetheus <3D35839A-7722-4740-B93D-51157F7D5E70>

inbox/queue/.extraction-debug/2026-03-08-intercept-openai-trust-us-surveillance.json

@@ -0,0 +1,36 @@
+{
+  "rejected_claims": [
+    {
+      "filename": "voluntary-safety-commitments-reduce-to-ceo-public-statements-without-external-enforcement.md",
+      "issues": [
+        "no_frontmatter"
+      ]
+    },
+    {
+      "filename": "contractual-safety-bans-versus-aspirational-constraints-reveals-moral-pragmatic-governance-tradeoff.md",
+      "issues": [
+        "missing_attribution_extractor"
+      ]
+    }
+  ],
+  "validation_stats": {
+    "total": 2,
+    "kept": 0,
+    "fixed": 6,
+    "rejected": 2,
+    "fixes_applied": [
+      "voluntary-safety-commitments-reduce-to-ceo-public-statements-without-external-enforcement.md:set_created:2026-03-28",
+      "voluntary-safety-commitments-reduce-to-ceo-public-statements-without-external-enforcement.md:stripped_wiki_link:voluntary-safety-pledges-cannot-survive-competitive-pressure",
+      "voluntary-safety-commitments-reduce-to-ceo-public-statements-without-external-enforcement.md:stripped_wiki_link:only-binding-regulation-with-enforcement-teeth-changes-front",
+      "contractual-safety-bans-versus-aspirational-constraints-reveals-moral-pragmatic-governance-tradeoff.md:set_created:2026-03-28",
+      "contractual-safety-bans-versus-aspirational-constraints-reveals-moral-pragmatic-governance-tradeoff.md:stripped_wiki_link:government-designation-of-safety-conscious-AI-labs-as-supply",
+      "contractual-safety-bans-versus-aspirational-constraints-reveals-moral-pragmatic-governance-tradeoff.md:stripped_wiki_link:Anthropics-RSP-rollback-under-commercial-pressure-is-the-fir"
+    ],
+    "rejections": [
+      "voluntary-safety-commitments-reduce-to-ceo-public-statements-without-external-enforcement.md:no_frontmatter",
+      "contractual-safety-bans-versus-aspirational-constraints-reveals-moral-pragmatic-governance-tradeoff.md:missing_attribution_extractor"
+    ]
+  },
+  "model": "anthropic/claude-sonnet-4.5",
+  "date": "2026-03-28"
+}

inbox/queue/2026-03-08-intercept-openai-trust-us-surveillance.md

@@ -7,9 +7,12 @@ date: 2026-03-08
 domain: ai-alignment
 secondary_domains: []
 format: article
-status: unprocessed
+status: enrichment
 priority: high
 tags: [OpenAI, autonomous-weapons, surveillance, trust-based-governance, voluntary-safety, self-attestation, governance-architecture, Sam-Altman, Pentagon-contract]
+processed_by: theseus
+processed_date: 2026-03-28
+extraction_model: "anthropic/claude-sonnet-4.5"
 ---
 
 ## Content
@@ -48,3 +51,11 @@ The Intercept framed the Anthropic/OpenAI divergence as: Anthropic pursued a mor
 PRIMARY CONNECTION: voluntary-pledges-fail-under-competition — "trust us" is the endpoint this claim describes; institutional-gap — the absence of external verification is the gap
 WHY ARCHIVED: Altman quote captures the self-attestation endpoint of voluntary governance; the Anthropic/OpenAI comparison is unusually explicit about the moral vs. pragmatic tradeoff
 EXTRACTION HINT: The claim should focus on governance architecture, not company ethics: voluntary safety commitments without external enforcement reduce to CEO public statements. The "trust us" quote is the evidence.
+
+
+## Key Facts
+- OpenAI's Pentagon contract amended language includes 'intentionally' qualifier for surveillance prohibition
+- OpenAI's Pentagon contract surveillance prohibition covers only U.S. persons and nationals, not non-US persons
+- OpenAI's Pentagon contract text remains non-public
+- Anthropic sought contractual bans on autonomous weapons targeting and mass surveillance in government contracts
+- OpenAI's Pentagon contract allows 'any lawful purpose' with added aspirational constraints