From Surveillance Capitalism to AI Imperialism | Chapter 07: Imperial control over access
The AI system is not simply a market of competing products. The United States has already moved from broad leadership rhetoric into formal doctrine, consultation processes, and implementation guidance. If frontier AI is governed through chokepoints, then intelligence itself becomes a stratified geopolitical infrastructure.
Imperial control over access; surveillance capitalism; AI imperialism; government capture; evidence-based dossier
How does imperial control over access function within the dossier’s larger argument, and what does the documentary record allow this chapter to establish?
This paper examines imperial control over access within the broader dossier From Surveillance Capitalism to AI Imperialism. It is intentionally written to circulate on its own: necessary context is restated, evidence boundaries are made explicit, and cited material is reproduced in paper-local form rather than delegated to repo navigation. It contributes to the series as a paper that establishes chokepoints, strategic access, and state-company interfaces that justify anti-capture obligations. Adjacent reinforcement appears in Chapter 06, Chapter 08, Chapter 10, but those cross-references are supportive rather than required for basic comprehension.
The AI system is not simply a market of competing products. It is also an architecture of chokepoints. Export controls, cloud dependence, chip access, and defense-state adjacency increasingly define who can build, scale, or even reach frontier intelligence. The question is no longer only who invents. It is who gets permission. 1
This paper is derived from the chapter corpus for Chapter 07 and is grounded in the project’s structured research OS. It relies on source notes, extracted claims, entity profiles, and timeline events already normalized under the dossier’s evidence hierarchy. Documented facts are asserted directly where the record supports them; disputed matters are marked as such; interpretive claims are bounded explicitly; and speculative overreach is isolated in the evidence-boundary appendix.
This paper asks: How does imperial control over access function within the dossier’s larger argument, and what does the documentary record allow this chapter to establish? Its working answer is clear from the start: The AI system is not simply a market of competing products. It is also an architecture of chokepoints. Export controls, cloud dependence, chip access, and defense-state adjacency increasingly define who can build, scale, or even reach frontier intelligence. The question is no longer only who invents. It is who gets permission. 2 The objective is not only to recount developments, but to show why the subject of this chapter belongs inside a structural argument about extraction, enclosure, legitimacy, and reform.
The United States has already moved from broad leadership rhetoric into formal doctrine, consultation processes, and implementation guidance. Executive action, federal requests for information, and corporate positioning now form a visible policy sequence rather than a vague background condition. 3
BIS guidance gives the advanced-computing export regime an operational enforcement layer. The politics of AI access is therefore not merely metaphorical. It is already administrative, technical, and executable. 4
Hyperscaler dependence is also documented. At least one major frontier lab publicly tied the majority of its workloads and future development to a single cloud giant, which means frontier capability is already routed through a small number of infrastructural landlords. 5
The formal government-facing layer is now public as well. OpenAI’s
OpenAI for Government, Anthropic’s Claude Gov,
and the Palantir-AWS-Anthropic classified distribution lane show that
frontier access is organized not only through markets but through
accredited, restricted, and government-priority channels. 6
The personnel interface is explicit too. The Army’s
Detachment 201 demonstrates that the boundary between
senior frontier-tech leadership and military institutions has become
porous in public view. That does not prove improper command. It does
prove public institutional integration. 7
In-Q-Tel, Palantir, and the Thiel-linked network provide a serious documentary basis for talking about venture-state-security adjacency without drifting into mythology. The structure is the point: chips, clouds, accredited environments, contracts, and strategic interfaces remain highly concentrated even if some model knowledge diffuses. 8
Interpretation: this is where
Limit Government Capture by AI Companies begins to look
structurally necessary. Once public support, export-control protection,
privileged state access, and defense channels become material to
frontier advantage, the resulting intelligence cannot be defended as a
purely ordinary private market outcome. 9
This paper sits inside a coordinated 11-paper architecture. Its nearest companions are Chapter 06 (AI as a dual-use rhetorical weapon), Chapter 08 (Power network, legitimacy, and institutional capture), Chapter 10 (What to do). In that series logic, Chapter 07 establishes chokepoints, strategic access, and state-company interfaces that justify anti-capture obligations.
This chapter must not confuse chokepoints with total closure. The evidence supports concentrated leverage, privileged lanes, and formal state-company interfaces, not omnipotent control over all capability diffusion. 10
This chapter must not confuse chokepoints with total closure. The evidence supports concentrated leverage, privileged lanes, and formal state-company interfaces, not omnipotent control over all capability diffusion. 11
If frontier AI is governed through chokepoints, then intelligence itself becomes a stratified geopolitical infrastructure. Access is no longer merely a subscription question. It becomes a question of empire, supply chains, protected lanes, and permission. 12
If frontier AI is governed through chokepoints, then intelligence itself becomes a stratified geopolitical infrastructure. But chokepoints alone do not explain public acceptance. In the architecture of this series, Chapter 07 therefore functions as a self-contained argument while also advancing the cumulative path toward the dossier’s three reforms.
This paper is designed to circulate independently. Public notes point readers to the real external documents first, then to the paper-local appendix entry that explains why each source matters.
The underlying research OS distinguishes among
Documented Fact, Disputed Fact,
Hypothesis / Interpretation, and
Speculative Narrative Risk. In Chapter 07, those boundaries
remain visible in the prose, in the bibliography, and in the appendix
sections that summarize claims and caution points.
whitehouse-eo-14179-2025. Relevance: That AI dominance was
elevated to explicit state doctrine in 2025.bis-guidance-advanced-computing-d5-macau-2026. Relevance:
That there is a primary U.S. government record directly describing
continued license requirements for advanced computing exports linked to
Country Group D:5 and Macau.amazon-anthropic-strategic-collaboration-2023. Relevance:
That at least one major frontier lab publicly committed the majority of
its workloads and future model development to a hyperscaler cloud
provider.fr-ai-action-plan-rfi-2025. Relevance: That the state
created an official consultation pathway for shaping the 2025 AI
doctrine.openai-us-ai-action-plan-2025. Relevance: That OpenAI
presents itself as a policy actor seeking to co-design national AI
strategy.openai-for-government-2025. Relevance: That by June 16,
2025 OpenAI had formally reorganized and expanded its U.S. public-sector
work under an OpenAI for Government initiative rather than
treating government use as…anthropic-claude-gov-2025. Relevance: That by June 6, 2025
Anthropic had publicly launched a government-specific model line for
U.S. national security customers.palantir-anthropic-aws-gov-partnership-2024. Relevance:
That by November 7, 2024 Anthropic model access for U.S. intelligence
and defense agencies was being publicly distributed through a
Palantir-plus-AWS stack.army-detachment-201-2025. Relevance: That by June 13, 2025
the U.S. Army had created a formal structure to embed senior
commercial-tech executives in a military innovation role.iqt-homepage-2026. Relevance: That
there is a formal institutional channel connecting strategic technology
development with intelligence and defense needs.palantir-gotham-2026. Relevance: That Palantir openly
markets Gotham to defense agencies and the intelligence community.meta-proxy-2022-thiel-retirement. Relevance: That Peter
Thiel’s formal role at Facebook / Meta lasted well beyond the early
startup period and remained active into 2022.deepseek-r1-reinforcement-learning-2025. Relevance: That by
January 2025 a non-U.S. lab was publicly claiming frontier-adjacent
reasoning performance in a primary technical paper.Source citation: Removing
Barriers to American Leadership in Artificial Intelligence
Institution / author: The White House
Date: 2025-01-23 Tier / type: T1 /
executive_order Why it matters in this paper: That AI
dominance was elevated to explicit state doctrine in 2025.
Corpus companion entry: Source S-130 Audit
ID: whitehouse-eo-14179-2025
Source citation: Guidance
Regarding Enforcement of License Requirements for Advanced Computing
Items for Entities Headquartered in Country Group D:5 and Macau
Institution / author: Bureau of Industry and Security
Date: 2026-05-31 Tier / type: T1 /
guidance_pdf Why it matters in this paper: That there
is a primary U.S. government record directly describing continued
license requirements for advanced computing exports linked to Country
Group D:5 and Macau. Corpus companion entry: Source
S-037 Audit ID:
bis-guidance-advanced-computing-d5-macau-2026
Source citation: Amazon
and Anthropic Announce Strategic Collaboration to Advance Generative
AI Institution / author: Amazon
Date: 2023-09-25 Tier / type: T1 /
press_release Why it matters in this paper: That at
least one major frontier lab publicly committed the majority of its
workloads and future model development to a hyperscaler cloud provider.
Corpus companion entry: Source S-002 Audit
ID:
amazon-anthropic-strategic-collaboration-2023
Source citation: Request
for Information on the Development of an Artificial Intelligence (AI)
Action Plan Institution / author: Federal Register
Date: 2025-02-06 Tier / type: T1 /
federal_register_notice Why it matters in this paper:
That the state created an official consultation pathway for shaping the
2025 AI doctrine. Corpus companion entry: Source S-054
Audit ID: fr-ai-action-plan-rfi-2025
Source citation: OpenAI’s
proposals for the U.S. AI Action Plan Institution /
author: OpenAI Date: 2025-03-13 Tier /
type: T1 / policy_submission Why it matters in this
paper: That OpenAI presents itself as a policy actor seeking to
co-design national AI strategy. Corpus companion entry:
Source S-096 Audit ID:
openai-us-ai-action-plan-2025
Source citation: Introducing
OpenAI for Government Institution / author: OpenAI
Date: 2025-06-16 Tier / type: T1 /
release Why it matters in this paper: That by June 16,
2025 OpenAI had formally reorganized and expanded its U.S. public-sector
work under an OpenAI for Government initiative rather than
treating government use as… Corpus companion entry:
Source S-090 Audit ID:
openai-for-government-2025
Source citation: Claude
Gov models for U.S. national security customers Institution
/ author: Anthropic Date: 2025-06-06
Tier / type: T1 / release Why it matters in
this paper: That by June 6, 2025 Anthropic had publicly
launched a government-specific model line for U.S. national security
customers. Corpus companion entry: Source S-004
Audit ID: anthropic-claude-gov-2025
Source citation: Anthropic
and Palantir Partner to Bring Claude AI Models to AWS for U.S.
Government Intelligence and Defense Operations Institution /
author: Palantir Technologies Date: 2024-11-07
Tier / type: T1 / press_release Why it matters
in this paper: That by November 7, 2024 Anthropic model access
for U.S. intelligence and defense agencies was being publicly
distributed through a Palantir-plus-AWS stack. Corpus companion
entry: Source S-098 Audit ID:
palantir-anthropic-aws-gov-partnership-2024
Source citation: Army
Launches Detachment 201: Executive Innovation Corps to Drive Tech
Transformation Institution / author: U.S. Army
Public Affairs Date: 2025-06-13 Tier /
type: T1 / press_release Why it matters in this
paper: That by June 13, 2025 the U.S. Army had created a formal
structure to embed senior commercial-tech executives in a military
innovation role. Corpus companion entry: Source S-026
Audit ID: army-detachment-201-2025
Source citation: In-Q-Tel Institution /
author: In-Q-Tel Date: n.d. Tier /
type: T1 / organization_homepage Why it matters in this
paper: That there is a formal institutional channel connecting
strategic technology development with intelligence and defense needs.
Corpus companion entry: Source S-074 Audit
ID: iqt-homepage-2026
Source citation: Gotham
Institution / author: Palantir Date:
n.d. Tier / type: T1 / product_page Why it
matters in this paper: That Palantir openly markets Gotham to
defense agencies and the intelligence community. Corpus
companion entry: Source S-099 Audit ID:
palantir-gotham-2026
Source citation: Meta
Platforms, Inc. 2022 Definitive Proxy Statement Institution
/ author: SEC Date: 2022-04-08 Tier /
type: T1 / proxy_statement Why it matters in this
paper: That Peter Thiel’s formal role at Facebook / Meta lasted
well beyond the early startup period and remained active into 2022.
Corpus companion entry: Source S-080 Audit
ID: meta-proxy-2022-thiel-retirement
Source citation: DeepSeek-R1: Incentivizing
Reasoning Capability in LLMs via Reinforcement Learning
Institution / author: arXiv Date:
2025-01-22 Tier / type: T1 / paper Why it
matters in this paper: That by January 2025 a non-U.S. lab was
publicly claiming frontier-adjacent reasoning performance in a primary
technical paper. Corpus companion entry: Source S-048
Audit ID:
deepseek-r1-reinforcement-learning-2025
Claim statement: That at least one major frontier
lab publicly committed the majority of its workloads and future model
development to a hyperscaler cloud provider. Evidence
status: Documented Fact Source support: Appendix
A3 Corpus companion entry: Claim C-004
Audit Claim ID:
amazon-anthropic-strategic-collaboration-2023-07-export-controls-compute-defense-access-claim-01
Claim statement: That Chapter 07 can document
compute dependence and cloud concentration with a primary source rather
than only with abstract market language. Evidence
status: Documented Fact Source support: Appendix
A3 Corpus companion entry: Claim C-005
Audit Claim ID:
amazon-anthropic-strategic-collaboration-2023-07-export-controls-compute-defense-access-claim-02
Claim statement: That chip access, cloud
infrastructure, model distribution, and capital investment were being
bundled together in a single strategic relationship. Evidence
status: Documented Fact Source support: Appendix
A3 Corpus companion entry: Claim C-006
Audit Claim ID:
amazon-anthropic-strategic-collaboration-2023-07-export-controls-compute-defense-access-claim-03
Claim statement: That by June 6, 2025 Anthropic had
publicly launched a government-specific model line for U.S. national
security customers. Evidence status: Documented Fact
Source support: Appendix
A7 Corpus companion entry: Claim C-010
Audit Claim ID:
anthropic-claude-gov-2025-07-export-controls-compute-defense-access-claim-01
Claim statement: That Anthropic publicly described
those models as already deployed in high-level national-security
agencies and restricted to classified environments. Evidence
status: Documented Fact Source support: Appendix
A7 Corpus companion entry: Claim C-011
Audit Claim ID:
anthropic-claude-gov-2025-07-export-controls-compute-defense-access-claim-02
Claim statement: That the company openly framed
model adaptation for defense, intelligence, and cybersecurity use cases
as part of its government offering. Evidence status:
Documented Fact Source support: Appendix
A7 Corpus companion entry: Claim C-012
Audit Claim ID:
anthropic-claude-gov-2025-07-export-controls-compute-defense-access-claim-03
Claim statement: That by June 13, 2025 the U.S. Army
had created a formal structure to embed senior commercial-tech
executives in a military innovation role. Evidence
status: Documented Fact Source support: Appendix
A9 Corpus companion entry: Claim C-078
Audit Claim ID:
army-detachment-201-2025-07-export-controls-compute-defense-access-claim-01
Claim statement: That the initial cohort explicitly
included executives from Palantir and OpenAI, not only traditional
defense contractors. Evidence status: Documented Fact
Source support: Appendix
A9 Corpus companion entry: Claim C-079
Audit Claim ID:
army-detachment-201-2025-07-export-controls-compute-defense-access-claim-02
Claim statement: That Chapter 07 and Chapter 08 can
document a direct institutional bridge between frontier/commercial AI
leadership and military structures without relying on rumor or guilt by
association. Evidence status: Documented Fact
Source support: Appendix
A9 Corpus companion entry: Claim C-080
Audit Claim ID:
army-detachment-201-2025-07-export-controls-compute-defense-access-claim-03
Claim statement: That there is a primary U.S.
government record directly describing continued license requirements for
advanced computing exports linked to Country Group D:5 and Macau.
Evidence status: Documented Fact Source
support: Appendix
A2 Corpus companion entry: Claim C-111
Audit Claim ID:
bis-guidance-advanced-computing-d5-macau-2026-07-export-controls-compute-defense-access-claim-01
Claim statement: That Chapter 07 can document a
concrete implementation layer beneath abstract talk of export controls
and access politics. Evidence status: Documented Fact
Source support: Appendix
A2 Corpus companion entry: Claim C-112
Audit Claim ID:
bis-guidance-advanced-computing-d5-macau-2026-07-export-controls-compute-defense-access-claim-02
Claim statement: That the AI Diffusion Rule and the
earlier November 2023 controls should be treated as connected parts of a
continuing export-control regime rather than isolated headlines.
Evidence status: Documented Fact Source
support: Appendix
A2 Corpus companion entry: Claim C-113
Audit Claim ID:
bis-guidance-advanced-computing-d5-macau-2026-07-export-controls-compute-defense-access-claim-03
Claim statement: That by January 2025 a non-U.S. lab
was publicly claiming frontier-adjacent reasoning performance in a
primary technical paper. Evidence status: Documented
Fact Source support: Appendix
A13 Corpus companion entry: Claim C-146
Audit Claim ID:
deepseek-r1-reinforcement-learning-2025-06-ai-double-use-reasoning-deepseek-claim-01
Claim statement: That reasoning capability was not
confined entirely to closed U.S. product stacks and was partly diffusing
through open publication and model release. Evidence
status: Documented Fact Source support: Appendix
A13 Corpus companion entry: Claim C-147
Audit Claim ID:
deepseek-r1-reinforcement-learning-2025-06-ai-double-use-reasoning-deepseek-claim-02
Claim statement: That Chapters 06 and 07 can
contrast frontier-access politics with a documented pattern of partial
technical diffusion. Evidence status: Documented Fact
Source support: Appendix
A13 Corpus companion entry: Claim C-148
Audit Claim ID:
deepseek-r1-reinforcement-learning-2025-06-ai-double-use-reasoning-deepseek-claim-03
Claim statement: That the state created an official
consultation pathway for shaping the 2025 AI doctrine. Evidence
status: Documented Fact Source support: Appendix
A4 Corpus companion entry: Claim C-167
Audit Claim ID:
fr-ai-action-plan-rfi-2025-08-power-networks-legitimacy-capture-claim-01
Claim statement: That legitimacy in AI was being
organized through formal administrative processes, not only informal
lobbying. Evidence status: Documented Fact
Source support: Appendix
A4 Corpus companion entry: Claim C-168
Audit Claim ID:
fr-ai-action-plan-rfi-2025-08-power-networks-legitimacy-capture-claim-02
Claim statement: That Chapter 8 can map
institutional capture or influence as a documented procedural field.
Evidence status: Documented Fact Source
support: Appendix
A4 Corpus companion entry: Claim C-169
Audit Claim ID:
fr-ai-action-plan-rfi-2025-08-power-networks-legitimacy-capture-claim-03
Claim statement: That the language of AI dominance
did not remain only in a political order, but moved into an official
federal consultation process. Evidence status:
Documented Fact Source support: Appendix
A4 Corpus companion entry: Claim C-170
Audit Claim ID:
fr-ai-action-plan-rfi-2025-claim-01
Claim statement: That OSTP and federal agencies
activated a formal channel to gather input from industry, academia, and
other actors on U.S. AI strategy. Evidence status:
Documented Fact Source support: Appendix
A4 Corpus companion entry: Claim C-171
Audit Claim ID:
fr-ai-action-plan-rfi-2025-claim-02
Claim statement: That the dossier can document a
clear institutional sequence between the presidential order, the RFI,
and later corporate policy positioning. Evidence
status: Documented Fact Source support: Appendix
A4 Corpus companion entry: Claim C-172
Audit Claim ID:
fr-ai-action-plan-rfi-2025-claim-03
Claim statement: That there is a formal
institutional channel connecting strategic technology development with
intelligence and defense needs. Evidence status:
Documented Fact Source support: Appendix A10 Corpus companion
entry: Claim C-234 Audit Claim ID:
iqt-homepage-2026-08-power-networks-legitimacy-capture-claim-01
Claim statement: That the adjacency between frontier
innovation and the state apparatus can be documented from primary
sources. Evidence status: Documented Fact
Source support: Appendix A10 Corpus companion
entry: Claim C-235 Audit Claim ID:
iqt-homepage-2026-08-power-networks-legitimacy-capture-claim-02
Claim statement: That Chapter 8 can discuss
legitimacy networks without relying on conspiratorial inference.
Evidence status: Documented Fact Source
support: Appendix A10
Corpus companion entry: Claim C-236 Audit Claim
ID:
iqt-homepage-2026-08-power-networks-legitimacy-capture-claim-03
Claim statement: That there is an organization
officially dedicated to connecting technological innovation with
intelligence, defense, and national-security needs. Evidence
status: Documented Fact Source support: Appendix A10 Corpus companion
entry: Claim C-237 Audit Claim ID:
iqt-homepage-2026-claim-01
Claim statement: That adjacency between venture
capital, frontier tech, and the state apparatus is not speculation: in
the case of IQT, it is expressed by the organization itself.
Evidence status: Documented Fact Source
support: Appendix A10
Corpus companion entry: Claim C-238 Audit Claim
ID: iqt-homepage-2026-claim-02
Claim statement: That the dossier can discuss a
technology-transfer infrastructure toward the state without depending on
conspiratorial sources. Evidence status: Documented
Fact Source support: Appendix A10 Corpus companion
entry: Claim C-239 Audit Claim ID:
iqt-homepage-2026-claim-03
Claim statement: That Peter Thiel’s formal role at
Facebook / Meta lasted well beyond the early startup period and remained
active into 2022. Evidence status: Documented Fact
Source support: Appendix
A12 Corpus companion entry: Claim C-262
Audit Claim ID:
meta-proxy-2022-thiel-retirement-11-facebook-thiel-intelligence-adjacency-claim-01
Claim statement: That the company itself presented
Thiel’s board tenure as spanning from 2005 to the end of his term at the
2022 annual meeting. Evidence status: Documented Fact
Source support: Appendix
A12 Corpus companion entry: Claim C-263
Audit Claim ID:
meta-proxy-2022-thiel-retirement-11-facebook-thiel-intelligence-adjacency-claim-02
Claim statement: That Thiel still held
committee-level governance authority at Meta immediately before his
departure from the board. Evidence status: Documented
Fact Source support: Appendix
A12 Corpus companion entry: Claim C-264
Audit Claim ID:
meta-proxy-2022-thiel-retirement-11-facebook-thiel-intelligence-adjacency-claim-03
Claim statement: That continuity between the
Facebook surveillance era and the Meta rebrand era included continuity
in elite board membership. Evidence status: Documented
Fact Source support: Appendix
A12 Corpus companion entry: Claim C-265
Audit Claim ID:
meta-proxy-2022-thiel-retirement-11-facebook-thiel-intelligence-adjacency-claim-04
Claim statement: That by June 16, 2025 OpenAI had
formally reorganized and expanded its U.S. public-sector work under an
OpenAI for Government initiative rather than treating
government use as incidental. Evidence status:
Documented Fact Source support: Appendix A6
Corpus companion entry: Claim C-298 Audit Claim
ID:
openai-for-government-2025-07-export-controls-compute-defense-access-claim-01
Claim statement: That OpenAI publicly tied
government access to secure environments, limited national-security
custom models, and structured forward visibility into upcoming
capabilities. Evidence status: Documented Fact
Source support: Appendix A6
Corpus companion entry: Claim C-299 Audit Claim
ID:
openai-for-government-2025-07-export-controls-compute-defense-access-claim-02
Claim statement: That OpenAI’s own record documents
a formal DoD CDAO pilot with a $200 million
ceiling under this initiative. Evidence status:
Documented Fact Source support: Appendix A6
Corpus companion entry: Claim C-300 Audit Claim
ID:
openai-for-government-2025-07-export-controls-compute-defense-access-claim-03
Claim statement: That OpenAI presents itself as a
policy actor seeking to co-design national AI strategy. Evidence
status: Documented Fact Source support: Appendix
A5 Corpus companion entry: Claim C-319
Audit Claim ID:
openai-us-ai-action-plan-2025-08-power-networks-legitimacy-capture-claim-01
Claim statement: That the company publicly links its
commercial and technical agenda to national-security and
industrial-policy language. Evidence status: Documented
Fact Source support: Appendix
A5 Corpus companion entry: Claim C-320
Audit Claim ID:
openai-us-ai-action-plan-2025-08-power-networks-legitimacy-capture-claim-02
Claim statement: That Chapter 8 can treat lobbying
and doctrinal positioning as part of the legitimacy architecture of
frontier firms. Evidence status: Documented Fact
Source support: Appendix
A5 Corpus companion entry: Claim C-321
Audit Claim ID:
openai-us-ai-action-plan-2025-08-power-networks-legitimacy-capture-claim-03
Claim statement: That OpenAI officially defended a
public-policy agenda combining global commercial expansion, export
controls, and preservation of access to copyrighted material for
training. Evidence status: Documented Fact
Source support: Appendix
A5 Corpus companion entry: Claim C-322
Audit Claim ID:
openai-us-ai-action-plan-2025-claim-01
Claim statement: That the company presents itself
not only as a model provider, but as an actor seeking to co-design U.S.
national AI strategy. Evidence status: Documented Fact
Source support: Appendix
A5 Corpus companion entry: Claim C-323
Audit Claim ID:
openai-us-ai-action-plan-2025-claim-02
Claim statement: That the language of national
security, competition with the PRC, and government adoption is already
integrated into its official 2025 discourse. Evidence
status: Documented Fact Source support: Appendix
A5 Corpus companion entry: Claim C-324
Audit Claim ID:
openai-us-ai-action-plan-2025-claim-03
Claim statement: That by November 7, 2024 Anthropic
model access for U.S. intelligence and defense agencies was being
publicly distributed through a Palantir-plus-AWS stack. Evidence
status: Documented Fact Source support: Appendix
A8 Corpus companion entry: Claim C-328
Audit Claim ID:
palantir-anthropic-aws-gov-partnership-2024-07-export-controls-compute-defense-access-claim-01
Claim statement: That classified-environment
frontier-model access was tied to specific accredited infrastructure and
defense-cleared platform arrangements rather than open public
availability. Evidence status: Documented Fact
Source support: Appendix
A8 Corpus companion entry: Claim C-329
Audit Claim ID:
palantir-anthropic-aws-gov-partnership-2024-07-export-controls-compute-defense-access-claim-02
Claim statement: That Chapter 07 can document a
concrete distribution channel where model access, cloud concentration,
accreditation, and defense use cases are bundled together.
Evidence status: Documented Fact Source
support: Appendix
A8 Corpus companion entry: Claim C-330
Audit Claim ID:
palantir-anthropic-aws-gov-partnership-2024-07-export-controls-compute-defense-access-claim-03
Claim statement: That Palantir openly markets Gotham
to defense agencies and the intelligence community. Evidence
status: Documented Fact Source support: Appendix A11 Corpus companion
entry: Claim C-331 Audit Claim ID:
palantir-gotham-2026-11-facebook-thiel-intelligence-adjacency-claim-01
Claim statement: That Palantir’s public product
framing already includes operational AI, deployment at the edge, and
mission-critical decision support. Evidence status:
Documented Fact Source support: Appendix A11 Corpus companion
entry: Claim C-332 Audit Claim ID:
palantir-gotham-2026-11-facebook-thiel-intelligence-adjacency-claim-02
Claim statement: That Palantir’s position inside
security-state and defense workflows does not need to be inferred from
hostile or conspiratorial sources because the company states it
directly. Evidence status: Documented Fact
Source support: Appendix
A11 Corpus companion entry: Claim C-333
Audit Claim ID:
palantir-gotham-2026-11-facebook-thiel-intelligence-adjacency-claim-03
Claim statement: That AI dominance was elevated to
explicit state doctrine in 2025. Evidence status:
Documented Fact Source support: Appendix
A1 Corpus companion entry: Claim C-477
Audit Claim ID:
whitehouse-eo-14179-2025-08-power-networks-legitimacy-capture-claim-01
Claim statement: That deregulatory and
dominance-oriented state framing can serve as an external source of
legitimacy for frontier firms. Evidence status:
Documented Fact Source support: Appendix
A1 Corpus companion entry: Claim C-478
Audit Claim ID:
whitehouse-eo-14179-2025-08-power-networks-legitimacy-capture-claim-02
Claim statement: That Chapter 8 can ground the idea
of institutional capture or mutual reinforcement in official state
language rather than metaphor alone. Evidence status:
Documented Fact Source support: Appendix
A1 Corpus companion entry: Claim C-479
Audit Claim ID:
whitehouse-eo-14179-2025-08-power-networks-legitimacy-capture-claim-03
Claim statement: That competition over AI was
officially framed by the U.S. government in terms of global dominance,
economic competitiveness, and national security. Evidence
status: Documented Fact Source support: Appendix
A1 Corpus companion entry: Claim C-480
Audit Claim ID:
whitehouse-eo-14179-2025-claim-01
Claim statement: That there was an explicit turn
toward a more deregulatory line aimed at removing internal barriers for
U.S. actors. Evidence status: Documented Fact
Source support: Appendix
A1 Corpus companion entry: Claim C-481
Audit Claim ID:
whitehouse-eo-14179-2025-claim-02
Claim statement: That the 2025 state framework
directly connected AI policy, national security, and administrative
reordering. Evidence status: Documented Fact
Source support: Appendix
A1 Corpus companion entry: Claim C-482
Audit Claim ID:
whitehouse-eo-14179-2025-claim-03
Claim statement: That current official records show
frontier labs and adjacent platform firms building formal
government-facing channels through secure offerings,
classified-environment access, accredited cloud stacks, and even
military personnel embedding rather than engaging the state only through
lobbying or abstract policy rhetoric. Evidence status:
Hypothesis / Interpretation Source support: Appendix A6,
Appendix
A7, Appendix
A8, Appendix
A9 Corpus companion entry: Claim C-437
Audit Claim ID:
vector07-formal-government-interface-synthesis-claim-01
Claim statement: That firms benefiting from
export-control protection, strategic state access, or privileged public
coordination should face stronger transparency and public-obligation
duties than ordinary software vendors. Evidence status:
Hypothesis / Interpretation Source support: Appendix
A1, Appendix
A5, Appendix
A2 Corpus companion entry: Claim C-438
Audit Claim ID:
vector07-public-obligations-for-protected-firms-claim-01
Claim statement: That the
Facebook-Thiel-Palantir-In-Q-Tel record supports a bounded argument
about networked governance continuity between surveillance platforms,
venture capital, and security-state technology ecosystems.
Evidence status: Hypothesis / Interpretation
Source support: Appendix
A12, Appendix A10
Corpus companion entry: Claim C-459 Audit Claim
ID:
vector11-networked-adjacency-synthesis-claim-01
Claim statement: That the strongest reading of the
Facebook-Thiel-Palantir-In-Q-Tel record is a structured adjacency model
built through board power, venture sponsorship, and government
contracting rather than evidence of covert command. Evidence
status: Hypothesis / Interpretation Source
support: Appendix
A12, Appendix A10, Appendix A11 Corpus companion
entry: Claim C-461 Audit Claim ID:
vector11-structured-adjacency-not-command-claim-01
Claim statement: That open reasoning papers or
public model releases eliminate the politics of compute concentration,
distribution control, or institutional gatekeeping around frontier AI.
Evidence status: Speculative Narrative Risk
Source support: Appendix
A13 Corpus companion entry: Claim C-432
Audit Claim ID:
vector06-open-publication-solves-control-overclaim-claim-01
Claim statement: That export controls, hyperscaler
dependence, and defense adjacency prove total state ownership or
seamless command over every frontier-model training, deployment, or
access pathway. Evidence status: Speculative Narrative
Risk Source support: Appendix
A2, Appendix
A3 Corpus companion entry: Claim C-439
Audit Claim ID:
vector07-total-state-command-overclaim-claim-01
Event summary: Meta’s proxy states that Peter Thiel
will retire from the board after the annual meeting Event
date: 2022-04-08 Source support: Appendix
A12 Corpus companion entry: Event T-017
Audit Event ID:
event-meta-proxy-thiel-retirement-2022-04-08
Event summary: Amazon and Anthropic tie
frontier-model development to AWS cloud and chips Event
date: 2023-09-25 Source support: Appendix
A3 Corpus companion entry: Event T-030
Audit Event ID:
event-amazon-anthropic-collaboration-2023-09-25
Event summary: Palantir, Anthropic, and AWS announce
a classified-environment channel for Claude in U.S. defense and
intelligence work Event date: 2024-11-07 Source
support: Appendix
A8 Corpus companion entry: Event T-059
Audit Event ID:
event-palantir-anthropic-aws-government-channel-2024-11-07
Event summary: DeepSeek publishes its R1 reasoning
paper and open release claims Event date: 2025-01-22
Source support: Appendix
A13 Corpus companion entry: Event T-065
Audit Event ID:
event-deepseek-r1-2025-01-22
Event summary: The White House makes sustaining and
enhancing U.S. global AI dominance official policy Event
date: 2025-01-23 Source support: Appendix
A1 Corpus companion entry: Event T-066
Audit Event ID:
event-eo-14179-2025-01-23
Event summary: The AI Action Plan RFI is published
in the Federal Register Event date: 2025-02-06
Source support: Appendix
A4 Corpus companion entry: Event T-068
Audit Event ID:
event-fr-rfi-2025-02-06
Event summary: OpenAI publishes its proposals for
the U.S. AI Action Plan Event date: 2025-03-13
Source support: Appendix
A5 Corpus companion entry: Event T-072
Audit Event ID:
event-openai-action-plan-2025-03-13
Event summary: Anthropic launches Claude Gov for
U.S. national security customers Event date: 2025-06-06
Source support: Appendix
A7 Corpus companion entry: Event T-077
Audit Event ID:
event-anthropic-claude-gov-2025-06-06
Event summary: The U.S. Army launches Detachment 201
and commissions senior tech executives into a reserve innovation corps
Event date: 2025-06-13 Source support:
Appendix
A9 Corpus companion entry: Event T-078
Audit Event ID:
event-army-detachment-201-2025-06-13
Event summary: OpenAI launches OpenAI for Government
and announces a DoD CDAO pilot Event date: 2025-06-16
Source support: Appendix A6
Corpus companion entry: Event T-079 Audit Event
ID: event-openai-for-government-2025-06-16
Event summary: BIS reaffirms license requirements
for advanced computing items tied to D:5 and Macau entities
Event date: 2026-05-31 Source support:
Appendix
A2 Corpus companion entry: Event T-092
Audit Event ID:
event-bis-advanced-computing-guidance-2026-05-31
Entity type: cloud_provider Role in this
paper: Actor relevant to
07_export_controls_compute_defense_access. Relevant sources in
this paper: Appendix
A3 Corpus companion entry: Entity E-009
Audit Entity ID: org-aws
Entity type: company Role in this
paper: Actor relevant to
07_export_controls_compute_defense_access. Relevant sources in
this paper: Appendix
A3 Corpus companion entry: Entity E-006
Audit Entity ID: org-amazon
Entity type: company Role in this
paper: Actor relevant to 04_data_to_models_copyright_weights,
06_ai_double_use_reasoning_deepseek,
07_export_controls_compute_defense_access,
08_power_networks_legitimacy_capture and others. Relevant
sources in this paper: Appendix
A3, Appendix
A7, Appendix
A8 Corpus companion entry: Entity E-008
Audit Entity ID: org-anthropic
Entity type: company Role in this
paper: Actor relevant to 06_ai_double_use_reasoning_deepseek,
07_export_controls_compute_defense_access. Relevant sources in
this paper: Appendix
A13 Corpus companion entry: Entity E-013
Audit Entity ID: org-deepseek
Entity type: company Role in this
paper: Actor relevant to 01_surveillance_capitalism,
03_covid_acceleration_and_geopolitics,
04_data_to_models_copyright_weights,
11_facebook_thiel_intelligence_adjacency and others. Relevant
sources in this paper: Appendix
A12 Corpus companion entry: Entity E-015
Audit Entity ID: org-facebook
Entity type: company Role in this
paper: Actor relevant to 04_data_to_models_copyright_weights,
06_ai_double_use_reasoning_deepseek,
07_export_controls_compute_defense_access,
08_power_networks_legitimacy_capture and others. Relevant
sources in this paper: Appendix
A5, Appendix A6
Corpus companion entry: Entity E-027 Audit
Entity ID: org-openai
Entity type: company Role in this
paper: Actor relevant to
07_export_controls_compute_defense_access,
08_power_networks_legitimacy_capture,
11_facebook_thiel_intelligence_adjacency. Relevant sources in
this paper: Appendix A11, Appendix
A8, Appendix
A9 Corpus companion entry: Entity E-029
Audit Entity ID: org-palantir
Entity type: government Role in this
paper: Actor relevant to
07_export_controls_compute_defense_access. Relevant sources in
this paper: Appendix
A2 Corpus companion entry: Entity E-010
Audit Entity ID: org-bis
Entity type: government Role in this
paper: Actor relevant to
07_export_controls_compute_defense_access,
08_power_networks_legitimacy_capture. Relevant sources in this
paper: Appendix
A1, Appendix
A4 Corpus companion entry: Entity E-039
Audit Entity ID: org-white-house
Entity type: government Role in this
paper: Actor relevant to
07_export_controls_compute_defense_access,
08_power_networks_legitimacy_capture. Relevant sources in this
paper: Appendix A6
Corpus companion entry: Entity E-014 Audit
Entity ID: org-dod-cdao
Entity type: government Role in this
paper: Actor relevant to
07_export_controls_compute_defense_access,
08_power_networks_legitimacy_capture. Relevant sources in this
paper: Appendix
A9 Corpus companion entry: Entity E-035
Audit Entity ID: org-us-army
Entity type: government_agency Role in this
paper: Actor relevant to
11_facebook_thiel_intelligence_adjacency. Relevant sources in
this paper: Appendix A10
Corpus companion entry: Entity E-012 Audit
Entity ID: org-cia
Entity type: organization Role in this
paper: Actor relevant to 01_surveillance_capitalism,
08_power_networks_legitimacy_capture,
11_facebook_thiel_intelligence_adjacency. Relevant sources in
this paper: Appendix A10
Corpus companion entry: Entity E-022 Audit
Entity ID: org-iqt
Entity type: person Role in this
paper: Actor relevant to 01_surveillance_capitalism,
08_power_networks_legitimacy_capture, 09_future_if_unchecked,
11_facebook_thiel_intelligence_adjacency. Relevant sources in
this paper: Appendix
A12 Corpus companion entry: Entity E-058
Audit Entity ID: person-peter-thiel
Entity type: person Role in this
paper: Actor relevant to
07_export_controls_compute_defense_access,
08_power_networks_legitimacy_capture, 09_future_if_unchecked,
10_counterstrategy_response_agenda and others. Relevant sources
in this paper: Appendix
A5 Corpus companion entry: Entity E-059
Audit Entity ID: person-sam-altman
OpenAI for Government
initiative rather than treating government use as incidental.Sources: The White House, Removing Barriers to American Leaders… (2025-01-23; see Appendix A1); Bureau of Industry and Security, Guidance Regarding Enforcement of Lic… (2026-05-31; see Appendix A2); Amazon, Amazon and Anthropic Announce Strateg… (2023-09-25; see Appendix A3)↩︎
Sources: The White House, Removing Barriers to American Leaders… (2025-01-23; see Appendix A1); Bureau of Industry and Security, Guidance Regarding Enforcement of Lic… (2026-05-31; see Appendix A2); Amazon, Amazon and Anthropic Announce Strateg… (2023-09-25; see Appendix A3)↩︎
Sources: The White House, Removing Barriers to American Leaders… (2025-01-23; see Appendix A1); Federal Register, Request for Information on the Develo… (2025-02-06; see Appendix A4); OpenAI, OpenAI’s proposals for the U.S. AI Ac… (2025-03-13; see Appendix A5)↩︎
Sources: Bureau of Industry and Security, Guidance Regarding Enforcement of Lic… (2026-05-31; see Appendix A2)↩︎
Sources: Amazon, Amazon and Anthropic Announce Strateg… (2023-09-25; see Appendix A3)↩︎
Sources: OpenAI, Introducing OpenAI for Government (2025-06-16; see Appendix A6); Anthropic, Claude Gov models for U.S. national s… (2025-06-06; see Appendix A7); Palantir Technologies, Anthropic and Palantir Partner to Bri… (2024-11-07; see Appendix A8)↩︎
Sources: U.S. Army Public Affairs, Army Launches Detachment 201: Executi… (2025-06-13; see Appendix A9)↩︎
Sources: In-Q-Tel, In-Q-Tel (n.d.; see Appendix A10); Palantir, Gotham (n.d.; see Appendix A11); SEC, Meta Platforms, Inc. 2022 Definitive… (2022-04-08; see Appendix A12); arXiv, DeepSeek-R1: Incentivizing Reasoning… (2025-01-22; see Appendix A13)↩︎
Sources: OpenAI, Introducing OpenAI for Government (2025-06-16; see Appendix A6); Anthropic, Claude Gov models for U.S. national s… (2025-06-06; see Appendix A7); Bureau of Industry and Security, Guidance Regarding Enforcement of Lic… (2026-05-31; see Appendix A2)↩︎
Sources: arXiv, DeepSeek-R1: Incentivizing Reasoning… (2025-01-22; see Appendix A13); Amazon, Amazon and Anthropic Announce Strateg… (2023-09-25; see Appendix A3)↩︎
Sources: arXiv, DeepSeek-R1: Incentivizing Reasoning… (2025-01-22; see Appendix A13); Amazon, Amazon and Anthropic Announce Strateg… (2023-09-25; see Appendix A3)↩︎
Sources: The White House, Removing Barriers to American Leaders… (2025-01-23; see Appendix A1); Bureau of Industry and Security, Guidance Regarding Enforcement of Lic… (2026-05-31; see Appendix A2); OpenAI, Introducing OpenAI for Government (2025-06-16; see Appendix A6)↩︎