Pentagon “blacklists” AI giant Anthropic; key point of contention: nuclear war

```

The Trump administration has officially designated the AI startup Anthropic as a “national security supply chain risk,” completely banning its use in federal government and military applications. This unprecedented rupture stems from fundamental disagreements between the two parties over whether AI should intervene in deadly military decisions such as nuclear warfare.

U.S. Secretary of Defense Pete Hegseth announced that any contractor involved with the U.S. military must immediately cease all business activities with Anthropic, and existing services will be phased out within six months. President Trump subsequently publicly pressured the company, warning that if it does not cooperate during the transition period, it will face civil and criminal charges.

This move has sent shockwaves through Silicon Valley and reshaped the rules of collaboration between tech giants and the U.S. military. For Anthropic, valued at as much as $38 billion and previously awarded about $200 million in military contracts, this not only means direct commercial losses, but also sends a strong compliance risk signal to the entire AI industry, demonstrating that the government will not allow companies to erect technological barriers in key national defense fields.

As Anthropic faces expulsion, its competitors OpenAI and Musk’s xAI are moving quickly to fill the market vacuum. According to The Wall Street Journal, OpenAI CEO Altman has told staff that he is negotiating a new cooperation agreement with the Pentagon, attempting to find a balance between meeting military needs and maintaining the company’s safety baseline.

Nuclear crisis hypothesis ignites “control” dispute

The core conflict between the Pentagon and Anthropic lies in the authority to apply AI systems during extreme military operations. Last month, the two sides suffered a serious clash while discussing an extreme nuclear warfare scenario: if the U.S. were attacked by intercontinental ballistic missiles, could the military use Anthropic's Claude AI system to assist in interception?

Pentagon officials accused Anthropic CEO Dario Amodei of trying to exercise veto power over military decision-making. Emil Michael, Deputy Secretary of Defense for Research and Engineering, publicly criticized Anthropic for placing its own ideology above national security, emphasizing that large tech companies must not be allowed to decide the outcome of wars.

Anthropic strongly refuted the military’s accusations, calling the claim about intercepting missiles “completely wrong.” Dario Amodei’s statement stressed that current AI systems are not reliable enough to drive autonomous weapons, and this would put soldiers and civilians in danger. The company insists on its core principle, namely, refusing to use AI technology for autonomous lethal weapons and domestic mass surveillance.

OpenAI and xAI accelerate to fill the market vacuum

Anthropic’s exit directly changes the competitive landscape of the AI defense market. OpenAI is attempting to offer the military a compromise solution. Altman told staff that the government is willing to include OpenAI’s “red lines” in the contract, allowing the company to build its own “safety stack.” This means that if the AI refuses to execute instructions that violate safety principles, the government will not force it to comply.

Meanwhile, Musk's xAI and its chatbot Grok, which has fewer restrictions, has become another highly attractive choice for the Pentagon and has been approved for use in some top-secret operations. However, according to the Wall Street Journal citing informed sources, several federal agencies, including the U.S. General Services Administration (GSA), have issued serious warnings about Grok’s safety and reliability, worrying that its data bias and vulnerability to manipulation could pose systemic risks.

The tech industry faces compliance and investment risk reassessment

Bloomberg analysis points out that Trump’s decision will prompt tech companies investing billions in AI research to rethink the best way to sign contracts with the federal government. Anthropic was previously the only AI system able to operate in the Pentagon’s classified cloud environment, with its Claude Gov tool being favored by defense staff for its ease of use. The ban poses enormous risks for both government and enterprise.

What worries the market even more are potential government coercive measures. Pentagon officials previously hinted that they might use government powers to compel Anthropic to hand over its technology. Industry experts warn that such extreme policy options would have a chilling effect on the whole sector, making investors and companies fear that deep cooperation with the Department of Defense might ultimately mean losing control of their core technologies.

Risk Warnings and Legal DisclaimerThe market has risks, investment should be prudent. This article does not constitute personal investment advice, nor does it take into account individual users’ specific investment goals, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions in this article fit their particular circumstances. Investing based on this is at your own risk. ```