On February 27, Bloomberg reported that U.S. AI startup Anthropic rejected the latest proposal from the Department of Defense, which aimed to resolve the impasse caused by the company setting conditions for the military to use its AI software. The standoff has jeopardized Anthropic's defense projects for the government.
U.S. Department of Defense
An Anthropic spokesman said in a statement Thursday that the Pentagon's new language, proposed as a compromise, failed to meet the company's requirement to put in place key safeguards when the military uses its AI tools. Those safeguards include a ban on using its technology to conduct mass surveillance of Americans and a ban on its technology being used in fully autonomous weapons systems.
The Pentagon rejected the requests and gave the company a Friday deadline to accept the government terms or be classified as a supply chain risk, a move that could bar it from working with other defense contractors. U.S. officials said the military wants to be able to use the company's AI tools within legal limits but should not be subject to any restrictions from Anthropic.
"These threats do not change our position: we cannot in good conscience accept their demands," Anthropic CEO Dario Amodei said in a statement on Thursday.
Defense Department officials pushed back and demanded that Claude, one of the few AI tools approved for classified cloud work, be used without any restrictions from the company. The U.S. Department of Defense has also threatened to use the Cold War-era Defense Production Act to use the company's software even if Anthropi objects.
Pentagon spokesman Sean Parnell said earlier Thursday that the Pentagon was not interested in mass surveillance or developing "autonomous weapons that can operate without human intervention."
"We will never allow any company to dictate how we make warfighting decisions," Parnell posted on