Trump bans Anthropic over ethical restrictions on the company's military use of artificial intelligence

Donald Trump has ordered federal agencies to stop using Anthropic's AI after days of dispute between the company and the Pentagon, which demanded to use the artificial intelligence without the company's ethical restrictions. The Pentagon demanded that Anthropic allow the use of its technology to command missile launches, fly autonomous military drones, and conduct mass surveillance. Dario Amodei, the company's CEO, refused, citing ethical and risk concerns. Until now, Anthropic's technology was used in the Department of Defense's classified system. Late Friday, the CEO of OpenAI, owner of ChatGPT, announced that he had reached an agreement with the administration to apply its models to the network of classified material. When all public contracts with Anthropic were canceled on Friday, the Pentagon also announced that it would declare the company a supply chain risk. This declaration is significant for the artificial intelligence lab after the clash over the limits of its technology's application. On Thursday, Amodei issued a statement amidst the ongoing standoff with Defense Secretary Pete Hegseth: "In a small set of cases, we believe AI can undermine, rather than defend, democratic values." It added: "Some uses are also simply beyond the limits of what current technology can do safely and reliably. Two of these use cases have never been included in our contracts with the Department of War, and we believe they should not be included now," specifically citing the use of autonomous weapons and the OpenAI Agreement.

"Tonight we reached an agreement with the Department of Defense to deploy our models on its classified network," wrote Sam Altman, CEO of OpenAI, in X. "AI safety and the broad sharing of its benefits are at the core of our mission. Two of our most important safety principles are the prohibition of the mass production of autonomous weapons systems. The Department of Defense shares these principles, reflects them in law and policy, and we have incorporated them into our agreement," he explained, aware of the clash between Secretary of Defense Pete Hegseth and Amodei in recent days. Altman also stated that they are asking the Department of Defense to "offer the same terms to other AI companies."

Threat to Anthropic

In a post on Truth Social, Trump announced that the Department of Defense and other agencies using Anthropic products will have six months to phase out the company's products. The president threatened Anthropic that if it does not assist with the transition, he will use "the full power of the presidency to compel them to comply, with significant civil and criminal consequences."

The decision is an exemplary and extraordinary punishment by the Trump administration against one of the leading companies that has kept the country at the forefront of artificial intelligence applied to national security. Furthermore, Trump's threat would effectively turn the company into a kind of corporate pariah. It's worth noting that Alphabet, the parent company of Google and Amazon, is among Anthropic's financial investors. The break also sets a precedent whereby only US legislation would restrict how AI is deployed on the battlefield, and the companies developing it would have no say. Thus, the Pentagon would preserve its full defense flexibility and would not be limited by warnings from the technology's creators against using unreliable AI to power weapons. In a statement, Anthropic has already said it will challenge any risk designation by the federal government in court. "We believe this designation would be legally unfounded and would set a dangerous precedent for any American company doing business with the government," the company stated.