Anthropic has denied unrestricted AI access to the U.S. department of defense, stepping into a a high-stakes standoff with the Pentagon. The fate of this $200 million government contract for Anthropic is likely to be decided on Friday.
Anthropic’s concerns
Anthropic CEO Dario Amodel posted a detailed statement on his discussions with the Department of War (DOW) on Thursday.
“The Department of War has stated they will only contract with AI companies who accede to ‘any lawful use’. They have threatened to remove us from their systems if we maintain these safeguards,” the statement claimed.
Anthropic said that the government should not be allowed to weaponize AI just because it can buy it. The AI company has also argued against the development of AI-powered fully autonomous weapons that eliminate humans from the decision-making loop.
Anthropic has alleged that the Pentagon has threatened to take it out of its systems unless these safeguards are dissolved.
“They have also threatened to designate us a ‘supply chain risk’—a label reserved for US adversaries, never before applied to an American company—and to invoke the Defense Production Act to force the safeguards’ removal,” Amodei said. “These latter two threats are inherently contradictory: one labels us a security risk; the other labels Claude as essential to national security.”
Pentagon denies allegations
Pentagon spokesperson Sean Parnell pushed back on Anthropic’s concerns saying that the agency has no interest in deploying AI to conduct mass surveillance. He also dismissed Anthropic’s claims of the department of defense looking to use AI to automate weapons.
“This narrative is fake,” Parnell posted on X, re-sharing a post from Google DeepMind chief scientist Jeff Dean wherein he said mass surveillance is unlawful in the U.S.
As per Pernell, all that the Pentagon is asking for is to use Anthropic’s Claude LLM for lawful purposes.
“This is a simple, common-sense request that will prevent Anthropic from jeopardizing critical military operations and potentially putting our warfighters at risk. We will not let ANY company dictate the terms regarding how we make operational decisions,” he said.
The Pentagon has given Anthropic up until 5:01 pm ET (2:01 PM GST) to come to a decision on whether it wishes to work with the Pentagon.
“Otherwise, we will terminate our partnership with Anthropic and deem them a supply chain risk for DOW,” Pernell noted.
Anthropic is also working with other U.S. government agencies including the national laboratories, civilian agencies, and AI safety institute around research and AI training activities.