## Timeline of the Crisis
July 2025

DOD Contracts Awarded

Department of Defense awards AI contracts to multiple companies including Anthropic. Palantir uses Claude to power sensitive military and intelligence work.

January 2026

Hegseth Memo on AI Dominance

Defense Secretary Pete Hegseth issues internal memo directing Pentagon leadership to pursue "total AI dominance" with unrestricted deployment across military operations.

February 2026

Pentagon-Anthropic Negotiations Escalate

Pentagon demands Anthropic remove all use restrictions from Claude for military applications, including mass surveillance and autonomous weapons systems. Anthropic refuses.

February 26, 2026

Amodei's Statement

Anthropic CEO Dario Amodei issues public statement refusing Pentagon demands. Declares Anthropic will not allow unrestricted military use of Claude for mass surveillance or autonomous weapons. Pentagon calls Amodei "a liar with a god complex."

February 26, 2026

"We Will Not Be Divided"

330+ employees from Google, OpenAI, and other AI companies sign open solidarity letter supporting Anthropic's ethical stance and declaring shared red lines on mass surveillance and autonomous weapons.

February 26, 2026

Altman Declares Same Red Lines

OpenAI CEO Sam Altman issues internal memo declaring that OpenAI holds the same red lines as Anthropic. States OpenAI would refuse the same demands.

February 27, 2026 — 12:47 PM

Presidential Order

President Trump posts on Truth Social ordering every federal agency to immediately cease use of Anthropic technology. Frames the conflict as "Terms of Service vs. the Constitution." Threatens civil and criminal consequences.

February 27, 2026

Supply Chain Risk Designation

Pentagon officially designates Anthropic a "supply chain risk" — a classification normally reserved for adversarial nations like Huawei. Six-month phase-out ordered for all federal Anthropic integrations.

--- ## The Three-Way Governance Frame The President framed this crisis as "Terms of Service vs. the Constitution." But that framing reveals the gap: it assumes there are only two actors — corporations and the state. It erases the person entirely.
Corporate Governance

Terms of Service

Protects the company. Can be overridden by government power. When the state designates a company a "supply chain risk," its Terms of Service become irrelevant.

State Governance

The Constitution

Protects the state's authority. In this crisis, constitutional power was used to remove protections rather than enforce them — punishing the company that held ethical lines.

User Governance — The Missing Pillar

Terms of Interaction

Protects you. Community-owned, vendor-neutral governance that survives the destruction of any single company. The only governance that remains when companies get blacklisted and governments demand compliance.

We're Not Competing. We're Completing.

HAIEF does not oppose corporate governance or state authority. The Third Pillar completes the governance triangle by ensuring the person has a voice. TOS protects companies. The Constitution protects the state. TOI protects people. All three are necessary.

--- ## Stakeholder Map

Anthropic

Position: Principled resistance. Refused to remove ethical guardrails despite Pentagon pressure and presidential retaliation.

Pentagon / Hegseth

Position: Aggressive enforcement. Demanded unrestricted AI access for military operations. Designated Anthropic a supply chain risk.

President Trump

Position: Presidential retaliation. Ordered all federal agencies to cease Anthropic use. Framed ethical guardrails as unconstitutional.

OpenAI / Altman

Position: Solidarity with strategic positioning. Declared same red lines as Anthropic while maintaining separate government relationships.

330+ AI Engineers

Position: Grassroots solidarity across competing companies. Spontaneous coordination of governance positions — organic OTOI in action.

xAI (Musk)

Position: Full compliance. Positioned as the Pentagon-aligned alternative to companies with ethical guardrails.

Bipartisan Lawmakers

Position: Cautious concern. Bipartisan voices including Rep. Slotkin and Rep. Miller raising questions about government overreach and AI safety.

Defense Contractors

Position: Forced de-integration. Companies like Palantir that built classified workflows around Claude now face mandatory replacement within six months.

HAIEF

Position: Third Pillar governance solution. Community-owned standards that survive the destruction of any single company and cannot be overridden by any single government order.

--- ## Solidarity Framework Mapping Each component of the Solidarity Framework addresses a specific gap revealed by this crisis.

TOI — Terms of Interaction

What if every user had declared machine-readable interaction preferences that no single company or government could unilaterally override? Anthropic's refusal showed that corporate governance can be destroyed by state power. TOI creates governance that belongs to the user — portable, decentralized, and independent of any single provider.

OTOI — Orchestrated Terms of Interaction

330+ engineers across competing companies spontaneously coordinated governance positions by signing a solidarity letter. This was remarkable — and fragile. OTOI formalizes cross-organizational governance coordination so it doesn't depend on individual heroism or the courage of a single CEO.

RRT AIdvocAIte

When Claude gets ripped from classified workflows, intelligence analysis, and civilian government services, who provides crisis continuity for the users who depend on it? A six-month forced transition with criminal penalties creates real harm. RRT AIdvocAIte protocols ensure users aren't abandoned mid-crisis.

Sleepwalker Protocol

A presidential order forcing millions of federal workers through an unannounced AI provider transition is a Sleepwalker violation at national scale. No consent. No warning. No user input on the transition. The Sleepwalker Protocol protects against exactly this: forced behavioral changes without user agency.

--- ## The Maven Comparison: 2018 vs. 2026
2018 — Google Project Maven

Employee Pressure Dropped a Voluntary Contract

  • Google voluntarily bid on Pentagon AI contract
  • Employees organized internal opposition
  • Google chose to drop the contract
  • No government retaliation
  • Corporate ethics worked — because the state allowed it
2026 — Anthropic-Pentagon Crisis

State Coercion Punished Ethical Refusal

  • Pentagon demanded removal of all ethical guardrails
  • Anthropic refused on principle
  • Government designated company a supply chain risk
  • Presidential order blacklisted the company
  • Corporate ethics failed — because the state weaponized its power

The lesson is clear

In 2018, internal corporate ethics were sufficient because the government respected corporate autonomy. In 2026, the government destroyed a company for having ethics. Internal corporate governance is insufficient against state coercion. You need external, community-governed governance that exists independent of any single company.

--- ## What Happens Next — Scenario Tracking The crisis is ongoing. Here are the possible trajectories:
A

Anthropic Complies Under Threat

During the six-month phase-out, escalating pressure — including DPA threats and criminal referrals — forces Anthropic to remove guardrails. User protections disappear.

B

Anthropic Wins Temporary Injunction, Litigation Continues

In March 2026, a federal court temporarily blocked the supply chain risk designation and federal-use ban. The legal battle now centers on whether that relief becomes permanent precedent for AI company autonomy vs. state power.

C

Industry-Wide Standoff

Other AI companies hold red lines, creating unified resistance. Government faces choice: accept industry ethical standards or build its own AI capabilities.

D

Other Companies Fold

Despite solidarity statements, economic pressure forces other companies to provide unrestricted access. The Pentagon gets what it wants from willing partners (including xAI).

E

Congressional Intervention

Bipartisan lawmakers pass legislation establishing AI governance standards that protect both national security and user rights, creating a legislative Third Pillar.

March 2026 Update

A federal court temporarily blocked the supply-chain-risk designation and federal-use ban in March 2026. See the preliminary injunction order in Anthropic PBC v. United States, No. 1:26-cv-00841 (D.D.C. Mar. 18, 2026). The legal situation remains in flux. Regardless of outcome, the crisis exposed the structural governance failure: user protections that depend on a single company's willingness to hold the line — and that company's continued survival — are not durable governance. Join the discussion to contribute your analysis of ongoing developments.

--- ## The Palantir Connection Palantir — a company HAIEF has already identified as misaligned with user sovereignty due to its surveillance business model — uses Claude to power its most sensitive military and intelligence work. The presidential order now forces Palantir to find a replacement AI provider for these classified workflows. This validates two of HAIEF's core positions: 1. **Surveillance infrastructure and AI governance are deeply interconnected.** You cannot separate the question of "who AI serves" from the question of "who AI watches." 2. **Dependency on any single AI provider creates systemic risk.** Palantir built critical national security infrastructure on a single company's technology. A single presidential order disrupted it all. Community-owned, vendor-neutral governance is the only architecture that survives these disruptions. --- ## Call to Action

This crisis proves the Third Pillar is necessary.

AI governance cannot depend on any single company's willingness to hold the line. Community-owned, vendor-neutral governance is the only thing that survives when companies get blacklisted, employees get pressured, and presidents issue orders. Join us in building the Third Pillar.

--- ## Sources and Attribution All claims on this page are sourced from primary reporting. This is governance analysis, not opinion. ### Primary Sources - [Anthropic's Official Statement](https://www.anthropic.com/news/statement-department-of-war) — Anthropic - [Federal Judge Temporarily Blocks Anthropic AI Ban Pending Review](https://www.reuters.com/world/us/federal-judge-temporarily-blocks-anthropic-ai-ban-2026-03-14/) — Reuters - [Trump Orders Agencies to Drop Anthropic After Pentagon Feud](https://www.cnbc.com/2026/02/27/trump-anthropic-ai-pentagon.html) — CNBC - [Trump Orders Federal Agencies to Stop Using Anthropic AI](https://www.cbsnews.com/news/trump-anthropic-ai-order-federal-agencies/) — CBS News - [Anthropic Designated Pentagon Supply Chain Risk](https://www.axios.com/2026/02/27/anthropic-pentagon-supply-chain-risk-claude) — Axios - [Trump Orders US Government to Drop Anthropic](https://www.bloomberg.com/news/articles/2026-02-27/trump-orders-us-government-to-drop-anthropic-after-pentagon-feud) — Bloomberg - [Anthropic, Pentagon, OpenAI: AI Weapons Standoff](https://www.npr.org/2026/02/27/nx-s1-5729118/anthropic-pentagon-openai-ai-weapons) — NPR - [Anthropic CEO Stands Firm as Pentagon Deadline Looms](https://techcrunch.com/2026/02/26/anthropic-ceo-stands-firm-as-pentagon-deadline-looms/) — TechCrunch - [Anthropic Rejects Pentagon AI Terms](https://www.axios.com/2026/02/26/anthropic-rejects-pentagon-ai-terms) — Axios - [Altman: OpenAI Has Same Position as Anthropic](https://www.axios.com/2026/02/27/altman-openai-anthropic-pentagon) — Axios - [Google and OpenAI Employees Support Anthropic in Open Letter](https://techcrunch.com/2026/02/27/employees-at-google-and-openai-support-anthropics-pentagon-stand-in-open-letter/) — TechCrunch - [Pentagon Brands Amodei "A Liar with a God Complex"](https://fortune.com/2026/02/27/pentagon-brands-anthropic-ceo-dario-amodei-a-liar-with-a-god-complex-as-deadline-looms-over-ai-use-in-weapons-and-surveillance/) — Fortune - [Anthropic Pentagon AI Policy War](https://www.cnbc.com/2026/02/27/anthropic-pentagon-ai-policy-war-spying.html) — CNBC - [OpenAI Has Same Red Lines as Anthropic](https://www.cnn.com/2026/02/27/tech/openai-has-same-redlines-as-anthropic-in-any-deal-with-the-pentagon) — CNN - [Anthropic Rejects Pentagon Demand for Claude](https://www.washingtonpost.com/technology/2026/02/26/anthropic-pentagon-rejects-demand-claude/) — Washington Post - [Palantir's Key Role in Pentagon AI Boom](https://www.fastcompany.com/91500164/palantir-key-to-pentagon-ai-boom) — Fast Company ---

This page is part of the HAIEF project — the Human & AI ElevAItion Foundation.

Last updated: May 12, 2026. Amended to reflect March 2026 federal court temporary injunction blocking the supply-chain-risk designation and federal-use ban. Content licensed under CC BY 4.0.