Entropic did not give in to the Pentagon’s demands; Declaration of opposition to the manufacture of automatic weapons

Less than 24 hours before the end of the Pentagon’s ultimatum, Anthropic did not give in to the demands of the US Department of Defense and officially rejected their request for unlimited military access to artificial intelligence. This bold decision was taken while other tech giants had already succumbed to pressure from the US government.

Anthropic executives rejected the military’s request for unrestricted access to its AI technology due to strong concerns about the use of AI for mass surveillance of American citizens and the creation of fully autonomous weapons without human supervision.

Dario Amodi, the CEO of Entropic, insisted on his company’s red lines after being summoned to the White House and meeting with Pete Hegst, the US Secretary of Defense. He recently wrote in his official statement:

“I believe deeply in the importance of using artificial intelligence to defend the United States and other democracies and defeat our authoritarian enemies … but in a few cases we believe that artificial intelligence can undermine rather than defend democratic values. “Some applications are simply beyond the scope of what today’s technology can do safely and reliably.”

Entropic’s non-acceptance of Pentagon requests

The US Department of Defense has threatened to list Entropic as a “supply chain threat” (commonly used for foreign adversaries) or use the Defense Production Act to force the company to remove security restrictions on its models. The Pentagon has asked all military contractors to assess their reliance on the cloud language model, so that it may label the US company as a supply chain risk.

In response to these pressures, Dario Amodi pointed out the obvious contradiction of the government and wrote:

“These two threats are inherently contradictory; One calls us a security risk and the other considers the cloud essential for national security. Regardless, these threats do not change our position; “We cannot in good conscience agree to their request.”

In his statement, Anthropic’s CEO confirmed that semi-automatic weapons are critical to defense and even fully automatic weapons may become necessary for national security in the future, but current technology is not yet ready for this. He says they will never provide the military with a product that endangers the lives of soldiers and civilians.

In the final part of his statement, Dario Amoudi calmly assured the army that their withdrawal will not affect military operations. He wrote about this:

“If the Department of Defense decides to terminate our relationship with Entropic, we will work with a smooth transition to another provider to avoid any disruption to military planning, operations or other critical missions.”

Competing companies such as OpenAI and xAI have embraced the Pentagon’s new requirements, but Anthropic is the only company that has military-ready classification models and still insists on its ethical principles.