WayToClawEarn
High impactThe Verge

U.S. Defense Secretary calls Anthropic CEO an "ideological lunatic": AI military cooperation is in stalemate

U.S. Secretary of Defense Pete Hegseth publicly called Anthropic CEO Dario Amodei an "ideological lunatic" during a Senate hearing because Anthropic refused to accept the U.S. Department of Defense's AI service terms. The incident revealed fundamental differences in the cooperation between AI companies and the military, which will have a profound impact on the application of AI agents and automation tools in the defense field.

WayToClawEarn EditorialPublished May 1, 2026Updated Aug 8, 2026

Editorial review of public sources · AI-assisted drafting. How we work · Original source

Core conclusion

On May 1, 2026, U.S. Secretary of Defense Pete Hegseth publicly called Anthropic CEO Dario Amodei an "ideological lunatic" at a Senate Armed Services Committee hearing. The comments stem from Anthropic's refusal to accept the U.S. Department of Defense's AI terms of service, amid fundamental differences between the two sides over the weaponized use of AI and the right of human oversight.

Key Points

  • Time of incident: May 1, 2026
  • Both parties involved: US Department of Defense (DoD) vs Anthropic
  • Core disagreement: Whether AI should be involved in military decision-making without human intervention
  • Scope of influence: AI Agent developers, automation tool manufacturers, defense contractors

Background and Hearing Site

During a question-and-answer session from Sen. Jacky Rosen (D-Nev.), she asked Hegseth about the Department of Defense’s contract dispute with Anthropic and whether Hegseth can guarantee that any targeting decisions made by the AI will have human oversight.

Hegseth's response was surprising - he focused his fire on Amodei personally, calling him an "ideological lunatic" and accusing Anthropic of refusing to "accept our terms of service." The language is extremely rare among senior U.S. officials’ public statements about tech CEOs.

Key Impact

DimensionsChangesWhat it means for the AI industryRecommended actions
Government RelationsAnthropic and DoD cooperation stalledAI security companies face the dilemma of "principles or orders"Develop an ethical framework for the military use of AI in advance
Industry reputationPublic personal attacks may affect government-enterprise relationsCommunication between technology companies and the government requires more professional PR strategiesEstablish a government affairs team and preset crisis communication plans
Policy trendsThe U.S. military may turn to more "compliant" AI suppliersAI companies with lax internal compliance may receive more defense ordersPay attention to the defense cooperation dynamics of Anthropic's competitors (such as OpenAI/Microsoft)
Developer EcosystemMilitarized applications of AI Agent trigger ethical discussionsOpen source AI Agent tools may face export control upgradesDevelopers using tools such as Claude Code need to pay attention to end-use terms

Deep analysis: Structural contradictions between AI companies and the military

This incident is not an isolated war of words, but a concentrated outbreak of structural conflicts between the AI industry and the military sector.

Safety Research: Anthropic has emphasized AI Safety (AI Safety) since its establishment, and its core mission is to ensure that AGI benefits mankind. CEO Dario Amodei has publicly stated multiple times that Anthropic will not use its most powerful models for military purposes.

Business Reality: The U.S. Department of Defense is one of the largest IT purchasers in the world, with an annual IT budget of over $50 billion. Defense contracts are a huge source of revenue for any AI company.

DIFFERENCE IN TERMS: The report mentioned that Anthropic "refused to accept our terms of service", implying that there are key differences between the two parties in the following aspects:

  • Can AI be used in autonomous weapons systems
  • Whether AI decision-making in military operations requires humans to "retain veto power"
  • Data sovereignty and audit rights of AI models in military scenarios

What it means for AI Agent developers

For developers building AI Agents using tools such as Claude, ChatGPT, Claude Code, etc., this incident has several immediate impacts:

1. Tightening of end-use terms: Anthropic may further strengthen its Acceptable Use Policy (AUP) to limit the application of its models in the defense field.

2. Differentiation of model authorization models: There may be a differentiation between "civilian version" and "defense version" AI models in the future. Different versions adopt different standards in terms of security and compliance.

3. Opportunities for Open Source Alternatives: Open source models such as DeepSeek V4 may find greater application in the defense technology sector because they are not bound by the terms of use of a single company.

4. Chain reaction of automation tools: Users of AI agents and automation frameworks such as n8n, OpenClaw, LangGraph, etc. need to pay attention to policy changes of their underlying model providers.

AI

Reference material

  • The Verge reported on May 1, 2026: Pete Hegseth calls Anthropic CEO Dario Amodei an "ideological lunatic"
  • Live transcript of U.S. Senate Armed Services Committee hearings
  • Anthropic Acceptable Use Policy (AUP) document

Tool entry (trigger tool floating card)

Related tools that appear naturally in the text: OpenAI, ChatGPT, Claude, Claude Code, DeepSeek, n8n, LangGraph, Hermes Agent, OpenClaw

Internal link guidance

View source →

Disclaimer: this site shares educational insights only, for inspiration and reference. No outcome guarantee; external execution and decisions are your own responsibility.