Anthropic Faced With DOD Ultimatum Over Government Usage

The Pentagon on Tuesday gave AI and safety research company Anthropic a 5:01 PM Friday deadline for allowing the US military unrestricted use of the technology or face a ban from all government contracts.

Claude Exclusive: Anthropic is the owner and developer of a family of large language models (LLM) called Claude and is currently the only AI model operating within the US government's classified systems. However, Elon Musk's xAI recently signed a deal to enter the classified systems, agreeing to the "all lawful use" standard that Anthropic is currently fighting against. Open AI and Google are also actively negotiating contracts with the government.

Supply Chain Risk: Secretary of Defense Pete Hegseth cited the Defense Production Act (DPA) of 1950 which grants a sitting president the authority to prioritize contracts and influence private companies' production to meet national defense requirements. As a part of the framework set forth in the act, companies can be designated as supply chain risks, classifying them as dangerous to US security and excluding them from any forms of government contracts or sub-contracts. This designation is often utilized for companies from adversarial nations, making the move unprecedented considering the company is based in the US.

Anthropic Reservations: Dario Amodei, CEO of Anthropic, is hesitant to allow unrestricted use over concerns of mass surveillance of Americans and use of the technology in autonomous weapons with no human oversight. An unnamed source told Just The News (Lean Right bias) that Amodei does not want Claude to be used for final targeting decisions in weapons systems, saying that "Claude is not immune to hallucinations and not reliable enough to avoid potentially lethal mistakes."

Idealistic Billionaire Owners: Associate professor of law at the University of Minnesota Law School Alan Rozenshtein called the threats to designate Anthropic a supply chain risk "completely inappropriate" and "deeply idiotic" in an interview with Bloomberg (Lean Left). However, Rozenshtein also conceded that the Pentagon has a point, saying that "defense officials don't want to be dictated to about how to handle potentially life-and-death battle scenarios, or depend on commercial software subject to the whims of idealistic billionaire owners." Hegseth emphasized Tuesday that when the government purchases Boeing planes, the company has no say in how the Pentagon uses the planes, saying the same should be true for Claude.

Lack Of Confidence: Some outlets on the left, like Bloomberg mentioned the slew of boat strikes on alleged drug traffickers, including the "double tap" strike on a boat in September, citing these as "one of the reasons companies such as Anthropic are forced to assert their own AI usage policy." The article cited various commentators who mentioned that under the Trump administration, companies cannot "have confidence its products will be used legally." 

Controller of Wartime Ethics: Some outlets on the right, like Breitbart (Right), suggested defense officials are caught between the choice of penalizing the company and recognizing their dependance on Claude. An opinion writer for the Washington Times (Lean Right) said "Silicon Valley executives seek control over lawful military missions," calling the rift between Anthropic and the Pentagon a "wound" and "a deeper dispute about who should decide the ethics of war and peace in the 21^st century."     

Written by the AllSides staff (of humans). Learn more. Support our mission. Suggest an improvement to this summary.