Anthropic said the quiet part out loud: ” No!”

On Tuesday, February 24, Department of War Pete Hegseth threatened the private AI company Anthropic with terminating its contract with the Pentagon if Anthropic did not acquiesce to Pentagon demands. At issue was Anthropic’s refusal to remove its restraint on using Anthropic’s products for mass domestic surveillance and military targeting decisions without human input. The deadline for acquiesce was Friday, February 27.

The threat to Anthropic went further. Non-compliance could result in Anthropic being labeled a supply chain risk, and the invocation of the Defense Production Act to force Anthropic’s compliance (an obviously illogical contradiction).

On February 26, Anthropic CEO Dario Amodei responded with an unequivocal “No.”

• AI makes it possible for government to assemble masses of scattered information and produce a comprehensive picture of anybody’s life. This use of AI is incompatible with democratic values.

• At present, AI systems are just not reliable enough to power autonomous weapons. Anthropic will not provide a product that will put American fighters and civilians at risk.

Dario Amodei further maintained that the law has not caught up with AI’s capabilities and potentials. Therefore, Pete Hegseth’s statement that the Pentagon will only do business with companies that accede to any lawful use of their products does not reflect reality.

So, Dario Amodei took a leap of faith. He made a public statement that he was saying “No”and why he was saying “No.” The kernel of his statement on this short sentence:

“Regardless, these threats do not change our position: we cannot in good conscience accede to their request.”

Predictably, all Hades broke loose, President Donald Trump directed all government agencies to stop using Anthropic’s AI systems, and Pete Hegseth proceeded in labeling Anthropic a supply chain risk to national security.

It will be interesting to see what AI company steps in to supply the federal government with AI capabilities without any red lines; like domestic surveillance of US citizens and unmanned military targeting totally fine – no worries.