The White House Explores Autonomous Weapons
The White House has initiated discussions around the potential use of autonomous weapons, creating uncertainty for companies developing artificial intelligence (AI) systems with Pentagon contracts. Anthropic, the creator of Claude—one of the most advanced chatbots available—has been the first to voice dissent. Their refusal to allow their technology to be utilized for lethal purposes has prompted employees at competing firms like Google and OpenAI to push back against such collaborations.
Threat from the Pentagon
Secretary of Defense Pete Hegseth has issued a deadline for Anthropic, the developer of Claude, to grant the Pentagon access to its AI for military applications by February 27. Failure to comply could result in the cancellation of their contracts, valued at $200 million, and label Anthropic as a “risk to the supply chain”—a designation previously assigned only to foreign firms. Anthropic's CEO, Dario Amodei, has firmly stated that the company cannot “in good faith” comply with the request, emphasizing that allowing their technology for autonomous weapons or mass surveillance contradicts democratic values.
Amodei expressed support for using AI to defend democracies and combat authoritarian regimes but criticized the Pentagon's threats to withdraw access to their technology as they maintain anti-military use safeguards. “These threats do not change our position,” he stated.
Internal Reactions in Silicon Valley
Amodei's stance has resonated within Silicon Valley. Following his public response, approximately 100 AI development employees at Google signed a letter addressed to Jeff Dean, the chief scientific officer of Google DeepMind. The letter articulated their concerns about Google's involvement with the Pentagon, cautioning against the use of Gemini for domestic surveillance or pilotless autonomous weapons. The employees urged management to prevent any agreements that would violate these principles, expressing a desire to maintain pride in their work.
Similarly, nearly 50 employees from OpenAI, along with 175 from Google, collaborated on an open letter condemning the Pentagon's negotiating tactics, which they describe as coercion. The letter calls for unity among engineers to reject the current demands from the Department of Defense, warning that the Pentagon aims to drive wedges between companies through fear of competition.
Contracts with the Pentagon
Since taking office, President Trump has emphasized the importance of AI supremacy, viewing it as a critical geostrategic asset. His administration has advocated for significant investments in AI, including initiatives like Stargate, which proposes a $500 billion allocation for AI development and military partnerships.
Major technology firms have become amenable to these military contracts. For instance, in February 2025, shortly after Trump's inauguration, Google lifted its ban on the development of weaponry and mass surveillance tools. Furthermore, Microsoft disclosed that it has been supplying advanced AI technology to the Israeli army since the Gaza conflict began, while OpenAI secured a $200 million contract in June 2025 to provide the Pentagon with generative AI solutions.
Anticipating these developments, OpenAI formed a partnership with Anduril—a military tech startup—in December 2024 to compete in defense contracts. Additionally, Meta announced its collaboration with military contractors Lockheed Martin and Booz Allen to leverage AI models for defense applications.
The ties between the tech industry and the Pentagon extend beyond financial agreements. In June of last year, the U.S. Army introduced the Innovation Executive Corps, appointing four lieutenant colonels from prominent tech firms, including executive members from Meta, OpenAI, and Palantir. This initiative aims to combine advanced technological skills with military innovation.
