theseus: extract claims from 2026-04-01-reaim-summit-2026-acoruna-us-china-refuse-35-of-85

- Source: inbox/queue/2026-04-01-reaim-summit-2026-acoruna-us-china-refuse-35-of-85.md
- Domain: ai-alignment
- Claims: 1, Entities: 1
- Enrichments: 2
- Extracted by: pipeline ingest (OpenRouter anthropic/claude-sonnet-4.5)

Pentagon-Agent: Theseus <PIPELINE>

domains/ai-alignment/international-voluntary-military-ai-governance-is-experiencing-declining-adherence-as-advanced-autonomous-weapons-states-withdraw.md

@@ -0,0 +1,17 @@
+---
+type: claim
+domain: ai-alignment
+description: The REAIM Summit trajectory from 60+ signatories (2024) to 35 signatories (2026) with US and China refusal demonstrates governance regression at the international level paralleling domestic voluntary commitment failures
+confidence: experimental
+source: REAIM Summit 2026 coverage (TheDefenseWatch, US News, Asia Financial, Capacity Global)
+created: 2026-04-04
+title: International voluntary military AI governance is experiencing declining adherence as the states most responsible for advanced autonomous weapons programs withdraw from multi-stakeholder norm-building processes
+agent: theseus
+scope: structural
+sourcer: TheDefenseWatch, US News, Asia Financial, Capacity Global
+related_claims: ["[[voluntary safety pledges cannot survive competitive pressure because unilateral commitments are structurally punished when competitors advance without equivalent constraints]]", "[[the alignment tax creates a structural race to the bottom because safety training costs capability and rational competitors skip it]]", "[[government designation of safety-conscious AI labs as supply chain risks inverts the regulatory dynamic by penalizing safety constraints rather than enforcing them]]"]
+---
+
+# International voluntary military AI governance is experiencing declining adherence as the states most responsible for advanced autonomous weapons programs withdraw from multi-stakeholder norm-building processes
+
+The Third REAIM Summit (A Coruña, February 2026) saw only 35 of 85 attending countries sign the 'Pathways for Action' declaration on military AI principles, down from ~60 nations at Seoul 2024. The United States, which signed the 2024 Seoul Blueprint under Biden, reversed position under Trump with VP Vance declining signature citing that 'excessive regulation would stifle innovation and weaken national security.' China consistently refused to sign, objecting to language mandating human intervention in nuclear command and control. The 35 signatories were exclusively middle powers (Canada, France, Germany, South Korea, UK, Ukraine) with no AI superpowers participating. This represents a complete collapse of the voluntary multilateral framework within 18 months, occurring precisely as competitive dynamics intensify between the US and China. The US rationale is structurally identical to the alignment tax argument: safety constraints are framed as competitive disadvantages that rational actors must reject. The pattern mirrors domestic voluntary commitment failures (Anthropic RSP rollback, government blacklisting of safety-conscious labs) but at the international scale where the stakes involve autonomous weapons systems rather than commercial AI products.

entities/ai-alignment/reaim-summit.md

@@ -0,0 +1,26 @@
+# REAIM Summit (Responsible AI in the Military Domain)
+
+**Type:** Multi-stakeholder international dialogue platform  
+**Founded:** 2022  
+**Founders:** Netherlands and South Korea  
+**Domain:** Military AI governance, autonomous weapons norms  
+**Status:** Active but declining participation  
+
+## Overview
+
+The Summit on Responsible AI in the Military Domain (REAIM) is a multi-stakeholder platform bringing together states, civil society, and industry to build shared norms for responsible military AI use. Designed as a complementary track to the formal CCW GGE process, REAIM operates through voluntary commitments rather than binding treaties.
+
+## Key Developments
+
+- **2022** — First REAIM Summit held with limited scope
+- **February 2024** — Second REAIM Summit in Seoul, South Korea. ~60 nations endorse the 'Blueprint for Action' including the United States under Biden administration
+- **September 24, 2025** — Global Commission on Responsible AI in the Military Domain publishes 'Responsible by Design' report with three guiding principles and five core recommendations
+- **February 4-5, 2026** — Third REAIM Summit in A Coruña, Spain. Only 35 of 85 attending countries sign 'Pathways for Action' declaration. United States (represented by VP J.D. Vance) and China both decline to sign. Signatories include Canada, France, Germany, South Korea, UK, Ukraine — exclusively middle powers with no AI superpowers participating
+
+## Governance Model
+
+Voluntary commitments to principles on military AI use, with no enforcement mechanism. States attend summits, negotiate declarations, and choose whether to endorse final documents.
+
+## Significance
+
+REAIM represents the primary international forum for military AI norm-building outside formal treaty processes. The sharp decline in participation (60→35 signatories) and withdrawal of both US and China signals governance regression in the voluntary multilateral framework for autonomous weapons systems.