The Pentagon revealed on Friday that it has inked deals with six of the world’s top artificial intelligence companies to use their software in the War Department’s classified work, after a deal with Anthropic for the use of its Claude systems fell through.
The department noted that it entered into agreements with SpaceX, OpenAI, NVIDIA, Reflection, Microsoft, and Amazon Web Services, as well as with Google, as reported earlier this week.
OpenAI and SpaceX had previously been deployed within the War Department for the tools to be used in military scenarios, but the agreement allows usage to expand to everyday operations.
Impact levels are security classifications for cloud computing systems under the Defense Information Systems Agency. Level six is reserved for the storage and processing of information classified up to the secret level. The next level up is not specified.
A Google spokesperson told the Washington Examiner on Tuesday that it is “committed to the private and public sector consensus that AI should not be used for domestic mass surveillance or autonomous weaponry without appropriate human oversight.”
Before the recent deals, Anthropic’s Claude systems were the only AI tools permitted in classified settings. But Anthropic’s CEO Dario Amodei raised concerns that the department may use the company’s software for autonomous warfare. War Secretary Pete Hegseth then labeled the company a “supply-chain risk,” resulting in several lawsuits between the two parties.
The Pentagon announced on Friday that it has inked deals with six of the world’s top artificial intelligence companies to use their software in the War Department’s classified work, after a deal with Anthropic for the use of its Claude systems fell through.
The department remarked that it entered into agreements with SpaceX, OpenAI, NVIDIA, Reflection, Microsoft, and Amazon Web Services, as well as with Google, as reported earlier this week.
“These agreements accelerate the transformation toward establishing the United States military as an AI-first fighting force and will strengthen our warfighters’ ability to maintain decision superiority across all domains of warfare,” the department’s release stated.
Integrating AI software into classified settings is intended to streamline workflow and “augment warfighter decision-making in complex operational environments,” and will be deployed to impact levels six and seven.
Impact levels are security classifications for cloud computing systems under the Defense Information Systems Agency. Level six is reserved for the storage and processing of information classified up to the secret level. The next level up is not specified.
GenAI.mil, the War Department’s AI platform, has already been widely deployed by personnel. The department remarked 1.3 million personnel have used the platform in its first five months to generate prompts and deploy thousands of autonomous agents.
Google’s deal with the department reportedly received some pushback from employees, particularly because the contract says that the company does not have “any right to control or veto lawful government operational decision-making.”
Prior to the agreement, more than 600 Google employees sent a letter to CEO Sundar Pichai, urging him not to agree to a deal that allows the company’s AI to be used for classified work. The employees noted that allowing the company’s software to be used in classified settings could lead to its deployment “in inhumane or extremely harmful ways.”
A Google spokesperson told the Washington Examiner on Tuesday that it is “committed to the private and public sector consensus that AI should not be used for domestic mass surveillance or autonomous weaponry without appropriate human oversight.”
Before the recent deals, Anthropic’s Claude systems were the only AI tools permitted in classified settings. But Anthropic’s CEO Dario Amodei raised concerns that the department may use the company’s software for autonomous warfare. War Secretary Pete Hegseth then labeled the company a “supply-chain risk,” resulting in several lawsuits between the two parties.
Despite the conflict with Anthropic and Google’s internal issues, the War Department emphasized that it plans to integrate AI as agency infrastructure.
“The Department will continue to build an architecture that prevents AI vendor lock and ensures long-term flexibility for the Joint Force,” the release remarked. “Access to a diverse suite of AI capabilities from across the resilient American technology stack will give warfighters the tools they need to act with confidence and safeguard the nation against any threat.”
