The US Department of War has reportedly raised new national security concerns about Anthropic. In a recent court filing tied to an ongoing legal dispute with the AI company, the Pentagon has cited the Claude developer’s use of foreign workers, including those from China. The Department of War’s filing highlights what it describes as another “problem” linked to the company, as it seeks to dismiss Anthropic’s lawsuit challenging its designation as a supply chain risk. The case underscores broader concerns within the US government about the AI industry’s reliance on global talent, particularly as Anthropic asks the court to overturn the designation, block its enforcement, and require federal agencies to withdraw directives to stop working with the company. The Pentagon’s latest concerns follow its designation of Anthropic as a supply chain risk earlier this month. The company is asking courts to overturn the designation, halt its enforcement, and require federal agencies to withdraw directives instructing them to drop the company.
What the Pentagon said about Anthropic hiring Chinese and other foreign workers
A recent declaration from Pentagon undersecretary Emil Michael (seen by Axios) states, “Anthropic employs a large number of foreign nationals to build and support its LLM products, including many from the People’s Republic of China (PRC).”Michael wrote that the use of those workers “increases the degree of adversarial risk should those employees comply with the PRC’s National Intelligence Law.”According to an Axios report, the filing adds that risks associated with other major American AI companies using foreign workers are mitigated by “the technical and security assurances of the other labs’ leadership, along with their consistently responsible and trustworthy behaviour” in their work with the Pentagon. “Anthropic’s case, however, is different,” Michael added.The Pentagon argues that its concerns about Anthropic extend beyond disagreements over domestic mass surveillance and autonomous weapons to broader national security risks. At the same time, it continues to rely on the company and remains open to extending the phasing-out deadline if needed.“As a matter of Department of War policy, we do not comment on ongoing litigation,” a Pentagon spokesperson said.Foreign-born workers reportedly account for a significant share of AI and tech talent in the US. Chinese-origin researchers accounted for roughly 38–40% of the top AI talent at US institutions as of 2023, according to a talent tracker, Axios notes.Anthropic was among the early adopters of operational security measures such as research compartmentalisation and audit trails, partly because it was the first AI lab to partner with the Pentagon. The company also said it disrupted an AI-driven cyber espionage campaign linked to China on its platform last year and restricted access from the PRC.“Insider threats are a genuine and tricky concern. Ironically, within the industry, Anthropic is widely considered to be the most serious and proactive about policing insider threats from foreign nationals and otherwise,” the Foundation for American Innovation’s Samuel Hammond said. A hearing on whether to grant Anthropic temporary relief is scheduled for March 24.
