Microsoft has filed a court brief in support of Anthropic’s legal challenge to the Department of Defense, supporting the AI company’s efforts to overturn an offensive designation that effectively bars it from government work.
Microsoft, which integrates Anthropic’s AI tools into systems it provides to the U.S. military, argued in a court brief filed this week in federal court in San Francisco that the temporary restraining order is needed to prevent serious disruption to suppliers whose products rely on its technology. Google, Amazon, Apple, and OpenAI also signed a letter of intent in support of Anthropic.
In a statement to the Guardian, Microsoft said: “The Department of the Army needs reliable access to the nation’s best technology, and everyone wants to ensure that AI is not used to start wars without mass domestic surveillance or human control. The government, the entire technology industry, and the American people need a path to achieving all of these goals together.”
Microsoft is one of the Pentagon’s most deeply integrated technology partners, holding a share of the Joe Biden-era military’s $9 billion joint combat cloud capabilities contract, along with Amazon, Google and Oracle, as well as billions more worth of individual software and enterprise services contracts. Microsoft’s government contracts span defense, intelligence and civilian agencies, and under the Trump administration, Microsoft signed a new multibillion-dollar contract in September to help advance advances in cloud services and AI in the federal government.
The lawsuit comes after Anthropic filed two lawsuits on Monday. One case in federal court in California and the other in the Washington, D.C. Circuit Court of Appeals challenged the Department of Defense’s decision to classify it as a supply chain risk, which has never before applied to U.S. companies.
The dispute stems from the collapse last month of contract negotiations over a $200 million contract to install Anthropic’s AI on sensitive military systems, just as the United States prepares for war against Iran.
Negotiations broke down after Anthropic insisted its technology should not be used for mass surveillance of Americans or to power autonomous lethal weapons, prompting Defense Secretary Pete Hegseth to call the company a supply chain risk. Last week, the Department of Defense formally notified Anthropic of the decision, and the company said government contracts had already begun to be canceled. On Thursday, Emile Michael, the Pentagon’s chief technology officer, told CNBC there was “no chance” the Pentagon would renegotiate with Anthropic after the designation.
In its complaint, Anthropic explained the limitations and hesitations behind its technology. “Antropic currently has no confidence that Claude will function reliably or safely, for example, in support of lethal autonomous warfare,” the filing states. “Thus, these usage restrictions are rooted in Anthropic’s unique understanding of Claude’s risks and limitations.”
The company also said its First Amendment rights were under attack and argued that the Pentagon used supply chain risk designations, typically given to companies with ties to foreign adversaries such as China, as ideological punishment for public positions on the safety of AI.
Meanwhile, a Pentagon investigation into the Tomahawk military attack on Shajara Tayebeh Elementary School, which reportedly killed at least 175 people, is underway, with a preliminary investigation reportedly finding Washington responsible for the killings, Iranian officials said. It is unclear whether AI was used in the airstrike, but a targeting error based on outdated data from the U.S. Defense Intelligence Agency appears to have been the cause, a person familiar with the investigation told The New York Times.
On Thursday, House Democrats sent a letter to the Pentagon expressing concern about the “broader pattern of civilian harm from U.S. and Israeli attacks on Iran” and asking for more information about the role of artificial intelligence in selecting targets.
“If artificial intelligence is used, is it subject to human review, and when?” the lawmakers wrote. “Was artificial intelligence, including the use of the Maven smart system, used to identify Shajara Tayebbe School as a target? If so, did a human verify the accuracy of this target?”
