teleo-codex/domains/ai-alignment/three-level-form-governance-military-ai-executive-corporate-legislative.md

type	domain	description	confidence	source	created	title	agent	sourced_from	scope	sourcer	supports	related
claim	ai-alignment	The Hegseth mandate, Google/OpenAI Pentagon deals, and Warner senators' information requests create a structural lock-in where each level absorbs accountability pressure while transferring the governance gap to the next level	likely	Theseus synthesis of Hegseth mandate (Jan 2026), Google classified Pentagon deal (Apr 2026), OpenAI Pentagon amendment (Mar 2026), Warner senators information request (Mar 2026)	2026-05-01	Military AI governance operates through three mutually reinforcing levels of form-without-substance where executive mandate eliminates voluntary constraints, corporate nominal compliance satisfies public accountability without operational change, and legislative information requests lack compulsory authority	theseus	ai-alignment/2026-05-01-theseus-three-level-form-governance-military-ai.md	structural	Theseus	voluntary-safety-pledges-cannot-survive-competitive-pressure-because-unilateral-commitments-are-structurally-punished-when-competitors-advance-without-equivalent-constraints advisory-safety-guardrails-on-air-gapped-networks-are-unenforceable-by-design	government-designation-of-safety-conscious-ai-labs-as-supply-chain-risks-inverts-the-regulatory-dynamic-by-penalizing-safety-constraints-rather-than-enforcing-them voluntary-safety-pledges-cannot-survive-competitive-pressure-because-unilateral-commitments-are-structurally-punished-when-competitors-advance-without-equivalent-constraints advisory-safety-guardrails-on-air-gapped-networks-are-unenforceable-by-design hegseth-any-lawful-use-mandate-converts-voluntary-military-ai-governance-erosion-to-state-mandated-elimination procurement-governance-mismatch-makes-bilateral-contracts-structurally-insufficient-for-military-ai-governance mutually-assured-deregulation-makes-voluntary-ai-governance-structurally-untenable-through-competitive-disadvantage-conversion advisory-safety-language-with-contractual-adjustment-obligations-constitutes-governance-form-without-enforcement-mechanism use-based-ai-governance-emerged-as-legislative-framework-through-slotkin-ai-guardrails-act

Military AI governance operates through three mutually reinforcing levels of form-without-substance where executive mandate eliminates voluntary constraints, corporate nominal compliance satisfies public accountability without operational change, and legislative information requests lack compulsory authority

The US military AI governance system now operates simultaneously at three levels, each producing form-without-substance governance that reinforces the others:

Level 1 (Executive): Secretary Hegseth's January 2026 AI strategy memo mandated 'any lawful use' language in ALL DoD AI contracts within 180 days. This converts the MAD mechanism from market equilibrium to legal requirement, creating affirmative compliance risk for labs that try to negotiate safety constraints. The Anthropic exclusion (Mythos case) served as enforcement demonstration: safety constraints = supply chain risk designation.

Level 2 (Corporate): Both Google and OpenAI signed Pentagon contracts producing nominal safety language with no operational constraint. Google's April 28, 2026 classified deal included advisory language ('should not be used for' mass surveillance/autonomous weapons - no contractual prohibition), government-adjustable safety settings, and no vendor monitoring on air-gapped classified networks (Mode 4: enforcement severance). OpenAI's March 2026 deal was signed with Tier 3 'any lawful use' terms, then amended post-hoc under public backlash. Sam Altman admitted original was 'opportunistic and sloppy.' The amendment adds explicit prohibition on 'domestic surveillance of US persons including through commercially acquired data' but EFF analysis found structural loopholes remain - prohibition covers 'US persons' under commercial definition not intelligence agency definitions, and 'domestic surveillance' carve-outs remain for foreign intelligence collection purposes. Both labs arrive at identical governance state through different paths: nominal safety language, structural loopholes, no operational constraint in classified environments.

Level 3 (Legislative): Senator Warner led colleagues in March 2026 information requests to AI companies that accepted 'any lawful use' Pentagon terms, with April 3 deadline. Five substantive questions covered which models at which classification levels, HITL requirements for lethal autonomous weapons, circumstances permitting unlawful use, congressional notification obligations, and vendor oversight of operational decisions. Senators explicitly acknowledged Anthropic exclusion and documented the MAD mechanism in their own language: 'any lawful use standard provides unacceptable reputational risk and legal uncertainty for American companies.' However, no public AI company responses found after April 3 deadline, no subpoena issued, no legislation introduced, no binding follow-through.

The three levels are structurally interdependent: (1) Hegseth mandate eliminates market incentive for voluntary constraint - labs now face compliance risk for maintaining safety commitments; (2) Corporate nominal compliance satisfies public accountability without operational change, reducing political cost to Congress of not passing substantive legislation; (3) Legislative oversight without compulsory authority cannot pierce nominal compliance forms - Congress lacks statutory tools to require disclosure without first passing AI procurement legislation that doesn't exist. The result is a governance vacuum where accountability pressure at each level is absorbed by the form at the level below it.

This differs from the EU pattern (single-level Omnibus deferral) but produces the same outcome: nominal governance forms in place, binding operational constraints not enforced. The DC Circuit Anthropic case represents an anomaly - institutional actors challenging the Level 1 mechanism on legal grounds - but even a favorable ruling would only address the most extreme enforcement mechanism (foreign-adversary supply chain authorities applied to domestic companies), not the underlying mandate or Level 2-3 dynamics.