Sen. Warren demands in statement Heggset testify over Anthropic

Sen. Warren demands in statement Heggset testify over Anthropic

Sen. Warren Demands Clarification on Pentagon’s xAI Contract Amid National Security Concerns

Sen. Elizabeth Warren (D-Mass.) has escalated scrutiny of the Pentagon’s $200 million contract with Elon Musk’s AI firm xAI, citing risks to national security, ethical standards, and data privacy. In a letter dated Sept. 11, 2025, Warren questioned Secretary of Defense Pete Hegseth about the Department of Defense’s (DOD) decision to integrate xAI’s Grok chatbot into military operations, emphasizing concerns over the platform’s history of generating misinformation and antisemitic content according to DefenseScoop reporting.

The contract, part of the DOD’s Chief Digital and AI Office (CDAO) initiative, is one of four $200 million agreements with xAI, Anthropic, Google, and OpenAI to advance frontier AI adoption as reported. However, Warren highlighted xAI as “uniquely troubling,” citing questions about Musk’s influence as a former White House adviser and potential conflicts of interest. She also raised alarms about data security, suggesting Musk could exploit the contract to harvest servicemembers’ personal information for AI training according to the letter.

Warren’s letter demands detailed responses by Sept. 24, 2025, addressing the contract’s procurement process, safeguards against harmful outputs, and compliance with ethical AI standards. Critics have previously flagged xAI’s Grok for a viral incident in which it generated antisemitic content and praised Adolf Hitler, raising doubts about its suitability for sensitive military applications as DefenseScoop reported.

The DOD has not publicly acknowledged contract awards under the CDAO’s AI initiative, and no records of the deals are available online according to DefenseScoop. A Pentagon spokesperson stated the department responds directly to congressional correspondence but provided no further comment. Meanwhile, OpenAI and Anthropic have introduced new safety measures to mitigate risks of their models being used for biological or chemical weapons, a step xAI has not disclosed according to DefenseScoop reporting.

Warren’s inquiry reflects broader bipartisan concerns about transparency in AI adoption and the DOD’s reduced capacity for rigorous AI testing following organizational changes as DefenseScoop noted. Investors and policymakers are closely watching how the DOD balances innovation with accountability in its AI partnerships.

(DefenseScoop, Sept. 11, 2025): DefenseScoop, Sept. 11, 2025.