Amazon Web Services, Microsoft and NVIDIA Secure Pentagon AI Contracts
Key Points
- AWS, Microsoft and NVIDIA signed contracts to supply AI tools for classified Pentagon networks.
- Reflection AI also joined the list of companies providing AI to the Defense Department.
- xAI, OpenAI and Google already have similar agreements with the Pentagon.
- Anthropic remains without a deal after refusing to drop safeguards on its Claude chatbot.
- Former President Trump ordered a six‑month halt on federal use of Anthropic products.
- OpenAI’s ChatGPT uninstall rates jumped 413% YoY in February following its defense contract.
- The deals aim to make the U.S. military an "AI‑first" fighting force, according to the Pentagon.
Amazon Web Services, Microsoft and NVIDIA have signed agreements to provide the U.S. Defense Department with artificial‑intelligence tools for use on classified networks, the Pentagon announced. The deals, which also include startup Reflection AI, join similar contracts already in place with xAI, OpenAI and Google. Anthropic remains the only major U.S. AI firm without a Pentagon agreement, after a dispute with the administration over safeguards on its Claude chatbot. The rapid expansion of AI in the military has sparked public backlash, as evidenced by a sharp rise in ChatGPT uninstall rates following OpenAI’s own deal.
The Pentagon confirmed on Tuesday that three of the nation’s biggest cloud and chip makers — Amazon Web Services, Microsoft and NVIDIA — have entered into contracts to supply artificial‑intelligence capabilities for "lawful operational use" on classified military networks. The agreements, announced alongside a similar pact with the AI startup Reflection AI, mark the latest wave of tech firms aligning their products with the Department of Defense’s push to embed AI across its warfighting architecture.
In a statement shared with Bloomberg, the Defense Department said the new contracts "accelerate the transformation toward establishing the United States military as an AI‑first fighting force." The move puts the three firms in the same circle as xAI, OpenAI and Google, which have already signed comparable deals. The only major U.S. AI provider left out of the roster is Anthropic, whose Claude chatbot remains unavailable to federal agencies.
Anthropic’s absence stems from a clash with the administration over the company’s built‑in safeguards. In February, Defense Secretary Pete Hegseth warned that Anthropic could be labeled a "supply chain risk" if it removed restrictions that prevent Claude from being used for mass surveillance or fully autonomous weapons. Anthropic refused to renegotiate, prompting former President Donald Trump to order all federal agencies to cease using Claude and other Anthropic products within six months. The disagreement has since spilled into a courtroom battle that continues to unfold.
Public reaction to the Pentagon’s AI procurement has been mixed at best. Market‑intelligence firm Sensor Tower reported that OpenAI saw uninstall rates for its ChatGPT app soar 413 percent year‑over‑year in February, shortly after the company announced its own defense contract. The spike suggests a growing unease among consumers about the intersection of commercial AI tools and military applications.
Industry analysts note that the rapid adoption of AI by the Defense Department reflects a broader trend: the government’s appetite for cutting‑edge technology is outpacing the sector’s traditional procurement cycles. Companies like AWS, Microsoft and NVIDIA are eager to lock in lucrative contracts, while critics warn that the speed of these deals leaves little room for thorough security reviews or public oversight.
For now, the Pentagon’s AI strategy appears to be moving forward on multiple fronts. With the addition of the three new partners, the department gains access to a broader suite of cloud‑based models, high‑performance GPUs and specialized AI services. Whether this expansion will translate into tangible battlefield advantages remains to be seen, but the partnership signals a clear intent to embed machine‑learning tools deep within the nation’s defense infrastructure.